{
  "niche": "small-business",
  "name": "AI Cost Per Task",
  "description": "Published AI list prices worked out per business task, in USD and AUD, with the source and the date each price was read.",
  "line": "The real cost of AI: published list prices per ordinary task, for owner-operators of small Australian businesses (roughly A$10 million in revenue).",
  "audience": "Owner-operators of small Australian businesses with roughly A$10 million in revenue.",
  "url": "https://aicostpertask.com",
  "readOn": "2026-10-20",
  "currencyNote": "costUsd is the calculated list-price total in US dollars. costAud converts it with the RBA rate below. Vendors do not publish the AUD amounts.",
  "exchangeRate": {
    "sourceName": "Reserve Bank of Australia",
    "sourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
    "rateDate": "2026-10-19",
    "readOn": "2026-10-20",
    "usdPerAud": "0.6975",
    "note": "The RBA table for 19 Oct 2026 lists the United States dollar at 0.6975 per Australian dollar. AUD figures on this site equal the USD figure divided by 0.6975 (multiplied by 10000/6975) and are rounded half up. The RBA page says the series should not be relied on for regulatory or commercial purposes. Vendors below do not publish these AUD amounts."
  },
  "tokenRule": {
    "sourceUrl": "https://help.openai.com/en/articles/4936856",
    "sourceReadOn": "2026-10-20",
    "statement": "OpenAI's token help article says that for English, 100 tokens are approximately 75 words."
  },
  "omitted": [
    "Monthly seat plans, including ChatGPT and Claude subscriptions, are not split into a per-task price. A per-task figure would require a task count the vendor does not publish.",
    "OpenAI image models are omitted. The image guide prices output in tokens and says the token count depends on quality and size. The pricing page does not fix a token count for GPT Image 2.5, so this sheet does not invent one.",
    "Transcription rows whose per-minute figure is in a column the OpenAI pricing page heads “Estimated cost” are omitted. GPT-Transcribe is included because its model page states a price per minute.",
    "Free tiers are not turned into a per-task price. The pages read here do not say how many of these scenarios a free allowance covers."
  ],
  "categories": [
    {
      "slug": "office",
      "name": "Office",
      "summary": "A customer email, a meeting transcript, an hour of audio, and pasted spreadsheet rows.",
      "tasks": [
        "draft-email",
        "summarise-meeting",
        "transcribe-hour",
        "spreadsheet-cleanup"
      ]
    },
    {
      "slug": "customers",
      "name": "Customers",
      "summary": "A written reply to one support ticket.",
      "tasks": [
        "support-ticket"
      ]
    },
    {
      "slug": "publishing",
      "name": "Publishing",
      "summary": "A blog post, one product image, and an 8-second video clip.",
      "tasks": [
        "blog-post",
        "product-image",
        "short-video"
      ]
    },
    {
      "slug": "documents",
      "name": "Documents",
      "summary": "Notes on a contract-length document. The row prices a model call, not a legal review.",
      "tasks": [
        "review-contract"
      ]
    },
    {
      "slug": "product",
      "name": "Product",
      "summary": "One request to code a small feature, with no agent loop.",
      "tasks": [
        "code-feature"
      ]
    }
  ],
  "labour": {
    "award": "Clerks—Private Sector Award 2020 [MA000002]",
    "classification": "Adult Level 2, Year 1, minimum hourly rate for ordinary hours (clause 16.1, Table 3)",
    "audPerHour": "29.45",
    "sourceName": "Fair Work Ombudsman, Clerks—Private Sector Award 2020",
    "sourceUrl": "https://awards.fairwork.gov.au/MA000002.html",
    "sourceReadOn": "2026-10-20",
    "effectiveFrom": "2026-07-01",
    "note": "The award page incorporates amendments up to and including 1 July 2026 (PR799283, operative from the first full pay period on or after 1 July 2026). Table 3 lists Level 2, Year 1 at $29.45 an hour and $1,119.10 a week. This is the minimum hourly rate for ordinary hours. It is not the casual rate ($36.81), not overtime, and not a loaded cost with superannuation, leave, or payroll tax."
  },
  "handNote": "handMinutes is an assumption for the by-hand comparison. It is not a time study. The hourly wage is the cited award rate.",
  "tasks": [
    {
      "slug": "draft-email",
      "category": "office",
      "name": "Draft a customer email",
      "summary": "Scenario: 75 words of instruction and facts in, 150 words of email out.",
      "handMinutes": 10
    },
    {
      "slug": "summarise-meeting",
      "category": "office",
      "name": "Summarise a meeting transcript",
      "summary": "Scenario, not a recording: 9,000 words of transcript in, 450 words of summary out.",
      "handMinutes": 45
    },
    {
      "slug": "transcribe-hour",
      "category": "office",
      "name": "Transcribe one hour of audio",
      "summary": "Sixty minutes of recorded audio, priced at the published per-minute rate.",
      "handMinutes": 240
    },
    {
      "slug": "spreadsheet-cleanup",
      "category": "office",
      "name": "Clean up spreadsheet data",
      "summary": "Scenario: 1,500 words of pasted rows in, 1,500 words of cleaned rows out. Not a spreadsheet product's monthly plan.",
      "handMinutes": 30
    },
    {
      "slug": "support-ticket",
      "category": "customers",
      "name": "Answer a support ticket",
      "summary": "Scenario: 225 words of ticket in, 150 words of reply out.",
      "handMinutes": 8
    },
    {
      "slug": "blog-post",
      "category": "publishing",
      "name": "Write a blog post",
      "summary": "Scenario: 150 words of brief in, 750 words of post out.",
      "handMinutes": 90
    },
    {
      "slug": "product-image",
      "category": "publishing",
      "name": "Generate a product image",
      "summary": "One 1024×1024 output image, using the token count printed on the pricing page. The text prompt is not included.",
      "handMinutes": 15
    },
    {
      "slug": "short-video",
      "category": "publishing",
      "name": "Generate a short video clip",
      "summary": "An 8-second clip. For Veo 3.1 that length is the length in Google's Veo guide. For Gemini Omni Flash it is this site's comparison length.",
      "handMinutes": 20
    },
    {
      "slug": "review-contract",
      "category": "documents",
      "name": "Review a contract-length document",
      "summary": "Scenario: 3,750 words of document in, 600 words of notes out. This prices a model call. It is not a legal review.",
      "handMinutes": 45
    },
    {
      "slug": "code-feature",
      "category": "product",
      "name": "Code a small feature",
      "summary": "Scenario: one request, 1,500 words of instruction and existing code in, 750 words of code and notes out. No agent loop.",
      "handMinutes": 120
    }
  ],
  "rows": [
    {
      "id": "draft-email--gpt-6-luna",
      "taskSlug": "draft-email",
      "taskName": "Draft a customer email",
      "category": "office",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00011000",
      "costAud": "0.00015771",
      "displayUsd": "US$0.00011",
      "displayAud": "A$0.000158",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario: 75 words of instruction and facts in, 150 words of email out. Input 75 words counted as 100 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(100 / 1,000,000) × US$0.10 + (200 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "draft-email--claude-haiku-5-5",
      "taskSlug": "draft-email",
      "taskName": "Draft a customer email",
      "category": "office",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00011000",
      "costAud": "0.00015771",
      "displayUsd": "US$0.00011",
      "displayAud": "A$0.000158",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario: 75 words of instruction and facts in, 150 words of email out. Input 75 words counted as 100 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(100 / 1,000,000) × US$0.10 + (200 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "draft-email--claude-sonnet-5-5",
      "taskSlug": "draft-email",
      "taskName": "Draft a customer email",
      "category": "office",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.00220000",
      "costAud": "0.00315412",
      "displayUsd": "US$0.0022",
      "displayAud": "A$0.003154",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario: 75 words of instruction and facts in, 150 words of email out. Input 75 words counted as 100 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(100 / 1,000,000) × US$2 + (200 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "draft-email--gemini-4-argon",
      "taskSlug": "draft-email",
      "taskName": "Draft a customer email",
      "category": "office",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.00220000",
      "costAud": "0.00315412",
      "displayUsd": "US$0.0022",
      "displayAud": "A$0.003154",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario: 75 words of instruction and facts in, 150 words of email out. Input 75 words counted as 100 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(100 / 1,000,000) × US$2.00 + (200 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "summarise-meeting--gpt-6-luna",
      "taskSlug": "summarise-meeting",
      "taskName": "Summarise a meeting transcript",
      "category": "office",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00150000",
      "costAud": "0.00215054",
      "displayUsd": "US$0.0015",
      "displayAud": "A$0.002151",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario, not a recording: 9,000 words of transcript in, 450 words of summary out. Input 9000 words counted as 12000 tokens. Output 450 words counted as 600 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(12000 / 1,000,000) × US$0.10 + (600 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "summarise-meeting--claude-haiku-5-5",
      "taskSlug": "summarise-meeting",
      "taskName": "Summarise a meeting transcript",
      "category": "office",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00150000",
      "costAud": "0.00215054",
      "displayUsd": "US$0.0015",
      "displayAud": "A$0.002151",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario, not a recording: 9,000 words of transcript in, 450 words of summary out. Input 9000 words counted as 12000 tokens. Output 450 words counted as 600 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(12000 / 1,000,000) × US$0.10 + (600 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "summarise-meeting--claude-sonnet-5-5",
      "taskSlug": "summarise-meeting",
      "taskName": "Summarise a meeting transcript",
      "category": "office",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.03000000",
      "costAud": "0.04301075",
      "displayUsd": "US$0.03",
      "displayAud": "A$0.043011",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario, not a recording: 9,000 words of transcript in, 450 words of summary out. Input 9000 words counted as 12000 tokens. Output 450 words counted as 600 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(12000 / 1,000,000) × US$2 + (600 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "summarise-meeting--gemini-4-argon",
      "taskSlug": "summarise-meeting",
      "taskName": "Summarise a meeting transcript",
      "category": "office",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.03000000",
      "costAud": "0.04301075",
      "displayUsd": "US$0.03",
      "displayAud": "A$0.043011",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario, not a recording: 9,000 words of transcript in, 450 words of summary out. Input 9000 words counted as 12000 tokens. Output 450 words counted as 600 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(12000 / 1,000,000) × US$2.00 + (600 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "transcribe-hour--gpt-transcribe",
      "taskSlug": "transcribe-hour",
      "taskName": "Transcribe one hour of audio",
      "category": "office",
      "vendor": "OpenAI",
      "model": "GPT-Transcribe",
      "modelId": "gpt-transcribe",
      "costUsd": "0.27000000",
      "costAud": "0.38709677",
      "displayUsd": "US$0.27",
      "displayAud": "A$0.387097",
      "pricingBasis": "Per-minute API list price",
      "assumptions": "Sixty minutes of recorded audio, priced at the published per-minute rate. 60 × US$0.0045. The model page prices transcription audio duration at US$0.0045 per minute. The file-transcription guide caps one file at 25 MB (https://developers.openai.com/api/docs/guides/speech-to-text), so an hour may need to be split or compressed. Splitting is not given a different per-minute price on that page.",
      "formula": "60 × US$0.0045",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/models/gpt-transcribe",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "spreadsheet-cleanup--gpt-6-luna",
      "taskSlug": "spreadsheet-cleanup",
      "taskName": "Clean up spreadsheet data",
      "category": "office",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00120000",
      "costAud": "0.00172043",
      "displayUsd": "US$0.0012",
      "displayAud": "A$0.00172",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario: 1,500 words of pasted rows in, 1,500 words of cleaned rows out. Not a spreadsheet product's monthly plan. Input 1500 words counted as 2000 tokens. Output 1500 words counted as 2000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(2000 / 1,000,000) × US$0.10 + (2000 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "spreadsheet-cleanup--claude-haiku-5-5",
      "taskSlug": "spreadsheet-cleanup",
      "taskName": "Clean up spreadsheet data",
      "category": "office",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00120000",
      "costAud": "0.00172043",
      "displayUsd": "US$0.0012",
      "displayAud": "A$0.00172",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario: 1,500 words of pasted rows in, 1,500 words of cleaned rows out. Not a spreadsheet product's monthly plan. Input 1500 words counted as 2000 tokens. Output 1500 words counted as 2000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(2000 / 1,000,000) × US$0.10 + (2000 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "spreadsheet-cleanup--claude-sonnet-5-5",
      "taskSlug": "spreadsheet-cleanup",
      "taskName": "Clean up spreadsheet data",
      "category": "office",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.02400000",
      "costAud": "0.03440860",
      "displayUsd": "US$0.024",
      "displayAud": "A$0.034409",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario: 1,500 words of pasted rows in, 1,500 words of cleaned rows out. Not a spreadsheet product's monthly plan. Input 1500 words counted as 2000 tokens. Output 1500 words counted as 2000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(2000 / 1,000,000) × US$2 + (2000 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "spreadsheet-cleanup--gemini-4-argon",
      "taskSlug": "spreadsheet-cleanup",
      "taskName": "Clean up spreadsheet data",
      "category": "office",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.02400000",
      "costAud": "0.03440860",
      "displayUsd": "US$0.024",
      "displayAud": "A$0.034409",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario: 1,500 words of pasted rows in, 1,500 words of cleaned rows out. Not a spreadsheet product's monthly plan. Input 1500 words counted as 2000 tokens. Output 1500 words counted as 2000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(2000 / 1,000,000) × US$2.00 + (2000 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "support-ticket--gpt-6-luna",
      "taskSlug": "support-ticket",
      "taskName": "Answer a support ticket",
      "category": "customers",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00013000",
      "costAud": "0.00018638",
      "displayUsd": "US$0.00013",
      "displayAud": "A$0.000186",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario: 225 words of ticket in, 150 words of reply out. Input 225 words counted as 300 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(300 / 1,000,000) × US$0.10 + (200 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "support-ticket--claude-haiku-5-5",
      "taskSlug": "support-ticket",
      "taskName": "Answer a support ticket",
      "category": "customers",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00013000",
      "costAud": "0.00018638",
      "displayUsd": "US$0.00013",
      "displayAud": "A$0.000186",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario: 225 words of ticket in, 150 words of reply out. Input 225 words counted as 300 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(300 / 1,000,000) × US$0.10 + (200 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "support-ticket--claude-sonnet-5-5",
      "taskSlug": "support-ticket",
      "taskName": "Answer a support ticket",
      "category": "customers",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.00260000",
      "costAud": "0.00372760",
      "displayUsd": "US$0.0026",
      "displayAud": "A$0.003728",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario: 225 words of ticket in, 150 words of reply out. Input 225 words counted as 300 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(300 / 1,000,000) × US$2 + (200 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "support-ticket--gemini-4-argon",
      "taskSlug": "support-ticket",
      "taskName": "Answer a support ticket",
      "category": "customers",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.00260000",
      "costAud": "0.00372760",
      "displayUsd": "US$0.0026",
      "displayAud": "A$0.003728",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario: 225 words of ticket in, 150 words of reply out. Input 225 words counted as 300 tokens. Output 150 words counted as 200 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(300 / 1,000,000) × US$2.00 + (200 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "blog-post--gpt-6-luna",
      "taskSlug": "blog-post",
      "taskName": "Write a blog post",
      "category": "publishing",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00052000",
      "costAud": "0.00074552",
      "displayUsd": "US$0.00052",
      "displayAud": "A$0.000746",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario: 150 words of brief in, 750 words of post out. Input 150 words counted as 200 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(200 / 1,000,000) × US$0.10 + (1000 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "blog-post--claude-haiku-5-5",
      "taskSlug": "blog-post",
      "taskName": "Write a blog post",
      "category": "publishing",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00052000",
      "costAud": "0.00074552",
      "displayUsd": "US$0.00052",
      "displayAud": "A$0.000746",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario: 150 words of brief in, 750 words of post out. Input 150 words counted as 200 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(200 / 1,000,000) × US$0.10 + (1000 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "blog-post--claude-sonnet-5-5",
      "taskSlug": "blog-post",
      "taskName": "Write a blog post",
      "category": "publishing",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.01040000",
      "costAud": "0.01491039",
      "displayUsd": "US$0.0104",
      "displayAud": "A$0.01491",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario: 150 words of brief in, 750 words of post out. Input 150 words counted as 200 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(200 / 1,000,000) × US$2 + (1000 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "blog-post--gemini-4-argon",
      "taskSlug": "blog-post",
      "taskName": "Write a blog post",
      "category": "publishing",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.01040000",
      "costAud": "0.01491039",
      "displayUsd": "US$0.0104",
      "displayAud": "A$0.01491",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario: 150 words of brief in, 750 words of post out. Input 150 words counted as 200 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(200 / 1,000,000) × US$2.00 + (1000 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "product-image--gemini-nano-banana-2-1",
      "taskSlug": "product-image",
      "taskName": "Generate a product image",
      "category": "publishing",
      "vendor": "Google",
      "model": "Gemini Nano Banana 2.1",
      "modelId": "gemini-nano-banana-2.1",
      "costUsd": "0.03360000",
      "costAud": "0.04817204",
      "displayUsd": "US$0.0336",
      "displayAud": "A$0.048172",
      "pricingBasis": "Per-image API list price (standard output image)",
      "assumptions": "One 1024×1024 output image, using the token count printed on the pricing page. The text prompt is not included. Standard image output is US$30 per 1M tokens. The pricing page says a 1024×1024 image consumes 1,120 tokens, which is US$0.0336. Prompt tokens are not included.",
      "formula": "1120 × US$30 / 1,000,000",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "product-image--gemini-3-1-flash-image",
      "taskSlug": "product-image",
      "taskName": "Generate a product image",
      "category": "publishing",
      "vendor": "Google",
      "model": "Gemini 3.1 Flash Image",
      "modelId": null,
      "costUsd": "0.06720000",
      "costAud": "0.09634409",
      "displayUsd": "US$0.0672",
      "displayAud": "A$0.096344",
      "pricingBasis": "Per-image API list price (standard output image)",
      "assumptions": "One 1024×1024 output image, using the token count printed on the pricing page. The text prompt is not included. The pricing page titles this model Gemini 3.1 Flash Image (Nano Banana 2) and does not print an API id in that heading. Standard image output is US$60 per 1M tokens. It says a 1024×1024 image consumes 1,120 tokens. 1120 × 60 / 1,000,000 = US$0.0672. The page rounds that equivalent to US$0.067. This row uses US$0.0672. Prompt tokens are not included.",
      "formula": "1120 × US$60 / 1,000,000",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "short-video--veo-3-1-lite",
      "taskSlug": "short-video",
      "taskName": "Generate a short video clip",
      "category": "publishing",
      "vendor": "Google",
      "model": "Veo 3.1 Lite",
      "modelId": "veo-3.1-lite-generate-preview",
      "costUsd": "0.40000000",
      "costAud": "0.57347670",
      "displayUsd": "US$0.40",
      "displayAud": "A$0.573477",
      "pricingBasis": "Per-second API list price (720p, with audio)",
      "assumptions": "An 8-second clip. For Veo 3.1 that length is the length in Google's Veo guide. For Gemini Omni Flash it is this site's comparison length. Veo 3.1 Lite 720p video with audio is US$0.05 per second on the pricing page. The Veo guide says Veo 3.1 generates 8-second videos (https://ai.google.dev/gemini-api/docs/veo). 8 × US$0.05 = US$0.40. The pricing page says the Veo 3.1 preview model ids, including veo-3.1-lite-generate-preview, are deprecated and shut down on 22 Oct 2026.",
      "formula": "8 × US$0.05",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "short-video--gemini-omni-flash",
      "taskSlug": "short-video",
      "taskName": "Generate a short video clip",
      "category": "publishing",
      "vendor": "Google",
      "model": "Gemini Omni Flash",
      "modelId": "gemini-omni-1.1-flash",
      "costUsd": "0.81088000",
      "costAud": "1.16255197",
      "displayUsd": "US$0.81088",
      "displayAud": "A$1.162552",
      "pricingBasis": "Per-second video output, from the published token rate",
      "assumptions": "An 8-second clip. For Veo 3.1 that length is the length in Google's Veo guide. For Gemini Omni Flash it is this site's comparison length. The pricing page says 720p video output is 5,792 tokens per second at US$17.50 per 1M tokens, which it calls approximately US$0.10 per second. This row is 8 seconds so it matches the Veo clip length. Eight seconds is this site's comparison length, not a length the Omni Flash section fixes. Input tokens are not included. Arithmetic: 5,792 × 8 × 17.50 / 1,000,000.",
      "formula": "5,792 × 8 × US$17.50 / 1,000,000",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "review-contract--gpt-6-luna",
      "taskSlug": "review-contract",
      "taskName": "Review a contract-length document",
      "category": "documents",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00090000",
      "costAud": "0.00129032",
      "displayUsd": "US$0.0009",
      "displayAud": "A$0.00129",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario: 3,750 words of document in, 600 words of notes out. This prices a model call. It is not a legal review. Input 3750 words counted as 5000 tokens. Output 600 words counted as 800 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(5000 / 1,000,000) × US$0.10 + (800 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "review-contract--claude-haiku-5-5",
      "taskSlug": "review-contract",
      "taskName": "Review a contract-length document",
      "category": "documents",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00090000",
      "costAud": "0.00129032",
      "displayUsd": "US$0.0009",
      "displayAud": "A$0.00129",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario: 3,750 words of document in, 600 words of notes out. This prices a model call. It is not a legal review. Input 3750 words counted as 5000 tokens. Output 600 words counted as 800 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(5000 / 1,000,000) × US$0.10 + (800 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "review-contract--claude-sonnet-5-5",
      "taskSlug": "review-contract",
      "taskName": "Review a contract-length document",
      "category": "documents",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.01800000",
      "costAud": "0.02580645",
      "displayUsd": "US$0.018",
      "displayAud": "A$0.025806",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario: 3,750 words of document in, 600 words of notes out. This prices a model call. It is not a legal review. Input 3750 words counted as 5000 tokens. Output 600 words counted as 800 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(5000 / 1,000,000) × US$2 + (800 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "review-contract--gemini-4-argon",
      "taskSlug": "review-contract",
      "taskName": "Review a contract-length document",
      "category": "documents",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.01800000",
      "costAud": "0.02580645",
      "displayUsd": "US$0.018",
      "displayAud": "A$0.025806",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario: 3,750 words of document in, 600 words of notes out. This prices a model call. It is not a legal review. Input 3750 words counted as 5000 tokens. Output 600 words counted as 800 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(5000 / 1,000,000) × US$2.00 + (800 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "code-feature--gpt-6-luna",
      "taskSlug": "code-feature",
      "taskName": "Code a small feature",
      "category": "product",
      "vendor": "OpenAI",
      "model": "GPT-6 Luna",
      "modelId": "gpt-6-luna",
      "costUsd": "0.00070000",
      "costAud": "0.00100358",
      "displayUsd": "US$0.0007",
      "displayAud": "A$0.001004",
      "pricingBasis": "Per-token API list price (standard, short context)",
      "assumptions": "Scenario: one request, 1,500 words of instruction and existing code in, 750 words of code and notes out. No agent loop. Input 1500 words counted as 2000 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Standard short-context list price on the OpenAI pricing page: US$0.10 input and US$0.50 output per 1M tokens. That page defines short context as 272,000 input tokens or fewer. No cached-input discount is applied.",
      "formula": "(2000 / 1,000,000) × US$0.10 + (1000 / 1,000,000) × US$0.50",
      "australianNotes": "OpenAI's data-controls guide lists Australia at au.api.openai.com with regional storage yes and regional processing no. It says non-US regions require abuse-monitoring approval and a modified-retention amendment, and that without regional processing OpenAI may process and temporarily store customer content outside the region. An asterisk on Australia says image support needs approval for enhanced zero data retention or enhanced modified abuse monitoring. The API pricing page says data-residency endpoints cost 10% more for eligible models released on or after 5 March 2026 (https://developers.openai.com/api/docs/pricing). The USD figure in this row is the standard list price, not that uplift. The pricing page does not mention AUD or GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://developers.openai.com/api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://developers.openai.com/api/docs/guides/your-data",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "code-feature--claude-haiku-5-5",
      "taskSlug": "code-feature",
      "taskName": "Code a small feature",
      "category": "product",
      "vendor": "Anthropic",
      "model": "Claude Haiku 5.5",
      "modelId": "claude-haiku-5-5",
      "costUsd": "0.00070000",
      "costAud": "0.00100358",
      "displayUsd": "US$0.0007",
      "displayAud": "A$0.001004",
      "pricingBasis": "Per-token API list price (prompt up to 100,000 tokens)",
      "assumptions": "Scenario: one request, 1,500 words of instruction and existing code in, 750 words of code and notes out. No agent loop. Input 1500 words counted as 2000 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Haiku 5.5 list price for a prompt up to 100,000 tokens: US$0.10 input and US$0.50 output per 1M tokens. This scenario is under that prompt length. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(2000 / 1,000,000) × US$0.10 + (1000 / 1,000,000) × US$0.50",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "code-feature--claude-sonnet-5-5",
      "taskSlug": "code-feature",
      "taskName": "Code a small feature",
      "category": "product",
      "vendor": "Anthropic",
      "model": "Claude Sonnet 5.5",
      "modelId": "claude-sonnet-5-5",
      "costUsd": "0.01400000",
      "costAud": "0.02007168",
      "displayUsd": "US$0.014",
      "displayAud": "A$0.020072",
      "pricingBasis": "Per-token API list price (standard global routing)",
      "assumptions": "Scenario: one request, 1,500 words of instruction and existing code in, 750 words of code and notes out. No agent loop. Input 1500 words counted as 2000 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Claude Sonnet 5.5 list price: US$2 input and US$10 output per 1M tokens. Global routing price, not the US-only 1.1× price. No cache discount.",
      "formula": "(2000 / 1,000,000) × US$2 + (1000 / 1,000,000) × US$10",
      "australianNotes": "Claude's data-residency page lists inference geo as global or us, and says the only workspace geo is us. It does not list Australia. The pricing page says US-only inference on Claude 4.6 and later is 1.1 times the token price, and that global routing uses the standard price (https://platform.claude.com/docs/en/about-claude/pricing). This row uses the standard price. That pricing page says all prices are in USD and does not mention GST. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://platform.claude.com/docs/en/about-claude/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://platform.claude.com/docs/en/manage-claude/data-residency",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    },
    {
      "id": "code-feature--gemini-4-argon",
      "taskSlug": "code-feature",
      "taskName": "Code a small feature",
      "category": "product",
      "vendor": "Google",
      "model": "Gemini 4 Argon",
      "modelId": "gemini-4-argon",
      "costUsd": "0.01400000",
      "costAud": "0.02007168",
      "displayUsd": "US$0.014",
      "displayAud": "A$0.020072",
      "pricingBasis": "Per-token API list price (standard paid tier, introductory rate)",
      "assumptions": "Scenario: one request, 1,500 words of instruction and existing code in, 750 words of code and notes out. No agent loop. Input 1500 words counted as 2000 tokens. Output 750 words counted as 1000 tokens. OpenAI's token help article says that for English, 100 tokens are approximately 75 words. The same count is used for every vendor. It is not that vendor's tokenizer. Not included: reasoning tokens, tool calls, retries, cache writes, and any second request. A vendor bill can be higher than this row. Gemini 4 Argon standard paid-tier list price as printed for 20 Oct 2026: US$2.00 input and US$10.00 output per 1M tokens through 30 Nov 2026, then US$4.00 and US$20.00 from 1 Dec 2026. This row uses the through-30-Nov figures. The page says the output price includes thinking tokens. This row still multiplies only the scenario's visible-text token count, because a thinking-token count for the scenario is not published.",
      "formula": "(2000 / 1,000,000) × US$2.00 + (1000 / 1,000,000) × US$10.00",
      "australianNotes": "The Gemini API available-regions page lists Australia among countries where the Gemini API and Google AI Studio are available. It does not describe an Australian storage or processing region. The pricing page labels paid prices as USD and does not mention GST. The AUD figure is this site's conversion, not a price Google shows in AUD. AUD in this row is USD divided by 0.6975, the RBA United States dollar rate per Australian dollar on 2026-10-19.",
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "sourceReadOn": "2026-10-20",
      "australianNotesSourceUrl": "https://ai.google.dev/gemini-api/docs/available-regions",
      "australianNotesReadOn": "2026-10-20",
      "usdPerAud": "0.6975",
      "exchangeRateDate": "2026-10-19",
      "exchangeRateSourceUrl": "https://www.rba.gov.au/statistics/frequency/exchange-rates.html",
      "exchangeRateReadOn": "2026-10-20"
    }
  ]
}