AI model cost calculator
Estimate what your chatbot, coding agent or RAG app will cost on GPT, Claude, Gemini, Grok, Mistral, DeepSeek and Meta models. Prices from official pages, last updated 2026-10-14.
Models
| Model | Provider | Per request | Per day | Per month | In / Out per 1M | Cached per 1M |
|---|
Sorted by cost, cheapest first. Excludes batch discounts, long-context surcharges, cache-write fees, tool fees and taxes.
AI API price table (per 1M tokens)
| Model | Provider | Input | Output | Cached | Context | Source |
|---|---|---|---|---|---|---|
| GPT-6 Astra | OpenAI | $10 | $50 | $1 | — | source · 2026-10-14 |
| GPT-6.1 Sol | OpenAI | $2 | $10 | $0.1 | — | source · 2026-10-14 |
| GPT-6 Luna | OpenAI | $0.1 | $0.5 | $0.01 | — | source · 2026-10-14 |
| GPT-5.6 Terra | OpenAI | $2 | $12 | $0.2 | — | source · 2026-10-14 |
| GPT-5.5 | OpenAI | $5 | $30 | $0.5 | — | source · 2026-10-14 |
| GPT-5.4 mini | OpenAI | $0.75 | $4.5 | $0.075 | — | source · 2026-10-14 |
| GPT-5 nano | OpenAI | $0.05 | $0.4 | $0.005 | — | source · 2026-10-14 |
| Claude Fable 5.1 | Anthropic | $10 | $50 | $0.25 | 1M | source · 2026-10-14 |
| Claude Opus 5.5 | Anthropic | $4 | $20 | $0.2 | 1M | source · 2026-10-14 |
| Claude Sonnet 5.5 | Anthropic | $2 | $10 | $0.1 | 1M | source · 2026-10-14 |
| Claude Sonnet 5 | Anthropic | $2 | $10 | $0.2 | 1M | source · 2026-10-14 |
| Claude Haiku 5.5 | Anthropic | $0.1 | $0.5 | $0.01 | — | source · 2026-10-14 |
| Claude Haiku 4.5 | Anthropic | $1 | $5 | $0.1 | — | source · 2026-10-14 |
| Gemini 4 Argon | $2 | $10 | $0.1 | — | source · 2026-10-14 | |
| Gemini 3.8 Flash | $0.75 | $3.75 | $0.075 | — | source · 2026-10-14 | |
| Gemini 3.1 Pro Preview | $2 | $12 | $0.2 | — | source · 2026-10-14 | |
| Gemini 3.5 Flash-Lite | $0.3 | $2.5 | $0.03 | — | source · 2026-10-14 | |
| Gemini 2.5 Flash-Lite | $0.1 | $0.4 | $0.01 | — | source · 2026-10-14 | |
| Grok 4.7 | xAI | $2 | $6 | $0.5 | 500K | source · 2026-10-14 |
| Grok 4.3 | xAI | $1.25 | $2.5 | $0.2 | 1M | source · 2026-10-14 |
| Grok Build 0.1 | xAI | $1 | $2 | $0.2 | 256K | source · 2026-10-14 |
| Mistral Large 4 | Mistral | $0.68 | $2.09 | $0.07 | 1M | source · 2026-10-14 |
| Mistral Medium 3.5 | Mistral | $1.5 | $7.5 | $0.15 | — | source · 2026-10-14 |
| Mistral Small 4 | Mistral | $0.15 | $0.6 | $0.015 | — | source · 2026-10-14 |
| Codestral | Mistral | $0.3 | $0.9 | $0.03 | — | source · 2026-10-14 |
| DeepSeek V4.1 Flash | DeepSeek | $0.3 | $1.2 | $0.006 | 1M | source · 2026-10-14 |
| DeepSeek V4 Pro | DeepSeek | $1.32 | $3.96 | $0.044 | 1M | source · 2026-10-14 |
| Muse Wonder 1.0 | Meta | $4 | $20 | $0.2 | — | source · 2026-10-14 |
| Muse Spark 1.3 | Meta | $1.25 | $4.25 | $0.15 | — | source · 2026-10-14 |
Also as prices.json and prices.md. Free to reuse with a link back.
One short email a week when model prices change. No spam, unsubscribe anytime.
FAQ
How is AI API cost calculated?
Cost = (input tokens × input price + output tokens × output price) ÷ 1,000,000. Cached input tokens are billed at the lower cached rate. TokenMeter multiplies per-request cost by requests per day, and by 30 for a monthly estimate.
How many tokens is a word?
For English text, one token is roughly three quarters of a word, so 1 word ≈ 1.33 tokens. Code and non-English text usually use more tokens. Tokenizers differ by model.
What is cached input?
Many providers discount input tokens that repeat a recently seen prompt prefix (system prompts, documents, agent context). Set the cached-input share to the fraction of input you expect to hit the cache.
Are these prices current?
Every price comes from the provider's official pricing page, with the source URL and last-verified date listed. Prices change often, so check the source before committing to a budget.
What isn't included?
Batch discounts, long-context surcharges, cache-write fees, tool calls (web search etc.), taxes, and regional uplifts are not included in the main estimate. Notes on each model page call these out.