GPT-6 Astra
Input / 1M$10.00
Cached input$1.00
Output / 1M$50.00
Example task$20.00
CHECKED 14 OCTOBER 2026
AI API pricing
What OpenAI, Anthropic, and Google charge for text, in US dollars. These are their published list rates, not a bill we measured.
OpenAIGPT-6 Astra$10.00 in · $50.00 out$20.001M in + 200k out
AnthropicClaude Opus 5.5$4.00 in · $20.00 out$8.001M in + 200k out
GoogleGemini 4 Argon$2.00 in · $10.00 out$4.001M in + 200k outThe example is one million input tokens and 200,000 output tokens, with no cache. A longer answer, a retry, or a tool call changes it. How to calculate AI API cost.
OpenAI
Standard rates for a prompt under 272,000 input tokens. Above that, OpenAI bills the whole request at twice the input and cache rates and 1.5 times the output rate. Batch and Flex are half of standard. Older models, including GPT-5.5, GPT-4o, and the o-series, are on OpenAI’s pricing page.
GPT-6.1 Sol
Input / 1M$2.00
Cached input$0.10
Output / 1M$10.00
Example task$4.00
GPT-6 Sol
Input / 1M$2.00
Cached input$0.20
Output / 1M$10.00
Example task$4.00
GPT-6 Luna
Input / 1M$0.10
Cached input$0.01
Output / 1M$0.50
Example task$0.20
GPT-5.6 Sol
Input / 1M$4.00
Cached input$0.40
Output / 1M$20.00
Example task$8.00
GPT-5.6 Terra
Input / 1M$2.00
Cached input$0.20
Output / 1M$12.00
Example task$4.40
GPT-5.6 Luna
Input / 1M$0.20
Cached input$0.02
Output / 1M$1.20
Example task$0.44
Global standard Claude API rates. US-only inference is 10 percent higher. Batch is half of standard. Cache writes cost more than a normal input; the table shows cache hits. Claude 4.7 and later use a tokenizer that Anthropic says produces about 30 percent more tokens for the same text.
Claude Fable 5.1
Input / 1M$10.00
Cached input$0.25
Output / 1M$50.00
Example task$20.00
Claude Mythos 5.1
Limited availability.
Input / 1M$10.00
Cached input$0.25
Output / 1M$50.00
Example task$20.00
Claude Opus 5.5
Input / 1M$4.00
Cached input$0.20
Output / 1M$20.00
Example task$8.00
Claude Sonnet 5.5
Input / 1M$2.00
Cached input$0.10
Output / 1M$10.00
Example task$4.00
Claude Haiku 5.5
Prompts up to 100,000 tokens.
Input / 1M$0.10
Cached input$0.01
Output / 1M$0.50
Example task$0.20
Claude Haiku 5.5
Prompts over 100,000 tokens.
Input / 1M$0.50
Cached input$0.05
Output / 1M$2.50
Example task$1.00
Claude Opus 5
Input / 1M$5.00
Cached input$0.50
Output / 1M$25.00
Example task$10.00
Claude Sonnet 5
Input / 1M$2.00
Cached input$0.20
Output / 1M$10.00
Example task$4.00
Claude Fable 5
Input / 1M$10.00
Cached input$1.00
Output / 1M$50.00
Example task$20.00
Claude Mythos 5
Limited availability.
Input / 1M$10.00
Cached input$1.00
Output / 1M$50.00
Example task$20.00
Earlier Anthropic models
Claude Opus 4.8
Input / 1M$5.00
Cached input$0.50
Output / 1M$25.00
Example task$10.00
Claude Opus 4.7
Input / 1M$5.00
Cached input$0.50
Output / 1M$25.00
Example task$10.00
Claude Opus 4.6
Input / 1M$5.00
Cached input$0.50
Output / 1M$25.00
Example task$10.00
Claude Opus 4.5
Listed for prompts up to 200,000 tokens.
Input / 1M$5.00
Cached input$0.50
Output / 1M$25.00
Example task$10.00
Claude Sonnet 4.6
Input / 1M$3.00
Cached input$0.30
Output / 1M$15.00
Example task$6.00
Claude Sonnet 4.5
Input / 1M$3.00
Cached input$0.30
Output / 1M$15.00
Example task$6.00
Claude Haiku 4.5
Input / 1M$1.00
Cached input$0.10
Output / 1M$5.00
Example task$2.00
Claude Opus 4.1
Retired, except on Bedrock and Google Cloud.
Input / 1M$15.00
Cached input$1.50
Output / 1M$75.00
Example task$30.00
Claude Opus 4
Retired, except on Google Cloud.
Input / 1M$15.00
Cached input$1.50
Output / 1M$75.00
Example task$30.00
Claude Sonnet 4
Retired, except on Bedrock and Google Cloud.
Input / 1M$3.00
Cached input$0.30
Output / 1M$15.00
Example task$6.00
Claude Haiku 3.5
Retired, except on Bedrock and Google Cloud.
Input / 1M$0.80
Cached input$0.08
Output / 1M$4.00
Example task$1.60
Paid Gemini API standard rates for text. Gemini output prices include thinking tokens. Batch is half of standard. Image, live, and audio models use different units and are on Google’s pricing page.
Gemini 4 Argon
Through 30 November 2026. From 1 December 2026, Google lists $4 input and $20 output. Not on the free tier.
Input / 1M$2.00
Cached input$0.10
Output / 1M$10.00
Example task$4.00
Gemini 3.1 Pro Preview
Prompts up to 200,000 tokens. Longer prompts are $4 input and $18 output.
Input / 1M$2.00
Cached input$0.20
Output / 1M$12.00
Example task$4.40
Gemini 3.8 Flash
Through 31 December 2026. From 1 January 2027, Google lists $1.50 input and $7.50 output.
Input / 1M$0.75
Cached input$0.075
Output / 1M$3.75
Example task$1.50
Gemini 3.6 Flash
Same introductory window as Gemini 3.8 Flash.
Input / 1M$0.75
Cached input$0.075
Output / 1M$3.75
Example task$1.50
Gemini 3 Flash Preview
Text, image, and video input. Audio input is $1.
Input / 1M$0.50
Cached input$0.05
Output / 1M$3.00
Example task$1.10
Gemini 3.1 Flash-Lite
Text, image, and video input. Audio input is $0.50.
Input / 1M$0.25
Cached input$0.025
Output / 1M$1.50
Example task$0.55
Gemini 2.5 Flash
Text, image, and video input. Audio input is $1.
Input / 1M$0.30
Cached input$0.03
Output / 1M$2.50
Example task$0.80