LLM API pricing comparison 2026
LLM API pricing for 36 current models from 5 providers (OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Mistral), in US dollars per 1 million tokens, standard tier. Use the calculator to see what your workload costs on each model, then check the provider pages for caching, batch and long prompt rules. Every price comes from the provider's own pricing page, linked below.
Cheapest model from each provider
Checked
| Provider | Input | Output |
|---|---|---|
| OpenAIGPT-6 Luna | $0.10 | $0.50 |
| Anthropic ClaudeClaude Haiku 5.5 | $0.10 | $0.50 |
| Google GeminiGemini 3.1 Flash-Lite | $0.25 | $1.50 |
| DeepSeekDeepSeek V4.1 Flash | $0.30 | $1.20 |
| MistralMinistral 3 3B | $0.10 | $0.10 |
US dollars per 1M tokens, standard tier. Two decimals from $0.10, two significant figures below.
LLM cost calculator
Set your tokens per request and monthly volume. Results sort cheapest first.
Includes reasoning or thinking tokens.
Cache write fees are not included.
Estimate tokens from text
About 0 tokens
A rough count at about 4 characters per token for English. Every provider tokenizes differently; Anthropic says Claude 4.7 and later produce about 30% more tokens for the same text.
| Model | Per request | Per month |
|---|---|---|
| Ministral 3 3BMistral · $0.10 in / $0.10 out | $0.00085 | $8.50 |
| GPT-6 LunaOpenAI · $0.10 in / $0.50 out | $0.0011 | $10.50 |
| Claude Haiku 5.5Anthropic · $0.10 in / $0.50 out | $0.0011 | $10.50 |
| Ministral 3 8BMistral · $0.15 in / $0.15 out | $0.0013 | $12.75 |
| Mistral Small 4Mistral · $0.15 in / $0.60 out | $0.0015 | $15.00 |
| Ministral 3 14BMistral · $0.20 in / $0.20 out | $0.0017 | $17.00 |
| GPT-5.6 LunaOpenAI · $0.20 in / $1.20 out | $0.0022 | $22.00 |
| GPT-5.4 nanoOpenAI · $0.20 in / $1.25 out | $0.0022 | $22.25 |
| Gemini 3.1 Flash-LiteGoogle · $0.25 in / $1.50 out | $0.0028 | $27.50 |
| CodestralMistral · $0.30 in / $0.90 out | $0.0028 | $28.50 |
| DeepSeek V4.1 FlashDeepSeek · $0.30 in / $1.20 out | $0.003 | $30.00 |
| Gemini 3.5 Flash-LiteGoogle · $0.30 in / $2.50 out | $0.0036 | $36.50 |
| Gemini 3 Flash PreviewGoogle · $0.50 in / $3.00 out | $0.0055 | $55.00 |
| Mistral Large 4Mistral · $0.70 in / $2.10 out | $0.0067 | $66.50 |
| Gemini 3.8 FlashGoogle · $0.75 in / $3.75 out | $0.0079 | $78.75 |
| Gemini 3.6 FlashGoogle · $0.75 in / $3.75 out | $0.0079 | $78.75 |
| GPT-5.4 miniOpenAI · $0.75 in / $4.50 out | $0.0083 | $82.50 |
| GPT-6.1 TerraOpenAI · $1.00 in / $5.00 out | $0.011 | $105.00 |
| Claude Haiku 4.5Anthropic · $1.00 in / $5.00 out | $0.011 | $105.00 |
| DeepSeek V4 ProDeepSeek · $1.32 in / $3.96 out | $0.013 | $125.40 |
| GPT-6.1 SolOpenAI · $2.00 in / $10.00 out | $0.021 | $210.00 |
| GPT-6 SolOpenAI · $2.00 in / $10.00 out | $0.021 | $210.00 |
| Claude Sonnet 5.5Anthropic · $2.00 in / $10.00 out | $0.021 | $210.00 |
| Claude Sonnet 5Anthropic · $2.00 in / $10.00 out | $0.021 | $210.00 |
| Gemini 4 ArgonGoogle · $2.00 in / $10.00 out | $0.021 | $210.00 |
| GPT-5.6 TerraOpenAI · $2.00 in / $12.00 out | $0.022 | $220.00 |
| Gemini 3.1 Pro PreviewGoogle · $2.00 in / $12.00 out | $0.022 | $220.00 |
| GPT-5.4OpenAI · $2.50 in / $15.00 out | $0.028 | $275.00 |
| GPT-5.6 SolOpenAI · $4.00 in / $20.00 out | $0.042 | $420.00 |
| Claude Opus 5.5Anthropic · $4.00 in / $20.00 out | $0.042 | $420.00 |
| Claude Opus 5Anthropic · $5.00 in / $25.00 out | $0.053 | $525.00 |
| GPT-5.5OpenAI · $5.00 in / $30.00 out | $0.055 | $550.00 |
| Claude Fable 5.2Anthropic · $8.00 in / $40.00 out | $0.084 | $840.00 |
| GPT-6 AstraOpenAI · $10.00 in / $50.00 out | $0.11 | $1,050.00 |
| Claude Fable 5.1Anthropic · $10.00 in / $50.00 out | $0.11 | $1,050.00 |
| GPT-5.5 ProOpenAI · $30.00 in / $180.00 out | $0.33 | $3,300.00 |
Standard tier prices per 1M tokens. Batch processing is about half price at OpenAI, Anthropic and Google. Taxes, tool fees and free tiers are not included.
OpenAI API pricing
OpenAI API pricing runs from $0.10 per million input tokens for GPT-6 Luna to $30 for GPT-5.5 Pro, with output tokens costing 5 to 6 times as much as input. Cached input is billed at a tenth of the input price or less on most models. Prices below are the standard tier for short prompts, from OpenAI's pricing page.
| Model | Input | Cached | Output | Context |
|---|---|---|---|---|
GPT-6 Astragpt-6-astraOver 272K tokens: $20.00 in, $75.00 out. | $10.00 | $1.00 | $50.00 | 1.05M |
GPT-6.1 Solgpt-6.1-solOver 272K tokens: $4.00 in, $15.00 out. | $2.00 | $0.10 | $10.00 | 1.05M |
GPT-6.1 Terragpt-6.1-terraOver 272K tokens: $2.00 in, $7.50 out. | $1.00 | $0.05 | $5.00 | 1.05M |
GPT-6 Solgpt-6-solOver 272K tokens: $4.00 in, $15.00 out. | $2.00 | $0.20 | $10.00 | Not listed |
GPT-6 Lunagpt-6-luna | $0.10 | $0.01 | $0.50 | 1.05M |
GPT-5.6 Solgpt-5.6-solOpenAI says its promotional pricing runs at least through 21 Nov 2026. | $4.00 | $0.40 | $20.00 | Not listed |
GPT-5.6 Terragpt-5.6-terraOver 272K tokens: $4.00 in, $18.00 out. Max input 922K tokens. | $2.00 | $0.20 | $12.00 | 1.05M |
GPT-5.6 Lunagpt-5.6-lunaOver 272K tokens: $0.40 in, $1.80 out. | $0.20 | $0.02 | $1.20 | Not listed |
GPT-5.5gpt-5.5Over 272K tokens: $10.00 in, $45.00 out. | $5.00 | $0.50 | $30.00 | Not listed |
GPT-5.5 Progpt-5.5-pro | $30.00 | Not listed | $180.00 | 1.05M |
GPT-5.4gpt-5.4Over 272K tokens: $5.00 in, $22.50 out. | $2.50 | $0.25 | $15.00 | 1.05M |
GPT-5.4 minigpt-5.4-miniMax input 272K tokens. | $0.75 | $0.075 | $4.50 | 400K |
GPT-5.4 nanogpt-5.4-nanoMax input 272K tokens. | $0.20 | $0.02 | $1.25 | 400K |
OpenAI pricing details, notes and price changes · Official source
Claude API pricing
Claude API pricing starts at $0.10 per million input tokens for Claude Haiku 5.5 and goes up to $10 for Claude Fable 5.1, with output at 5 times the input price. The newer Claude Fable 5.2 costs less, at $8 in and $40 out. Opus 5.5 is $4 in and $20 out, and Sonnet 5.5 is $2 in and $10 out. All five current models have a 1M token context window at standard pricing.
| Model | Input | Cached | Output | Context |
|---|---|---|---|---|
Claude Fable 5.2claude-fable-5-2 | $8.00 | $0.20 | $40.00 | 1M |
Claude Fable 5.1claude-fable-5-1 | $10.00 | $0.25 | $50.00 | 1M |
Claude Opus 5.5claude-opus-5-5 | $4.00 | $0.20 | $20.00 | 1M |
Claude Sonnet 5.5claude-sonnet-5-5 | $2.00 | $0.10 | $10.00 | 1M |
Claude Haiku 5.5claude-haiku-5-5Over 100K tokens: $0.50 in, $2.50 out. | $0.10 | $0.01 | $0.50 | 1M |
Claude Opus 5claude-opus-5 | $5.00 | $0.50 | $25.00 | Not listed |
Claude Sonnet 5claude-sonnet-5 | $2.00 | $0.20 | $10.00 | Not listed |
Claude Haiku 4.5claude-haiku-4-5 | $1.00 | $0.10 | $5.00 | Not listed |
Anthropic Claude pricing details, notes and price changes · Official source
Gemini API pricing
Gemini API pricing starts at $0.25 per million input tokens for Gemini 3.1 Flash-Lite, the cheapest model Google offers to new projects, and the Flash and Flash-Lite models also have a free tier. Gemini 4 Argon, the newest flagship, costs $2 in and $10 out until 30 November 2026, then doubles to $4 and $20. Output prices include thinking tokens.
| Model | Input | Cached | Output | Context |
|---|---|---|---|---|
Gemini 4 Argongemini-4-argonFrom December 1, 2026: $4.00 in, $20.00 out. Implicit caching only. | $2.00 | $0.10 | $10.00 | 1M |
Gemini 3.1 Pro Previewgemini-3.1-pro-previewOver 200K tokens: $4.00 in, $18.00 out. | $2.00 | $0.20 | $12.00 | Not listed |
Gemini 3.8 Flashgemini-3.8-flashFrom January 1, 2027: $1.50 in, $7.50 out. Free tier available. | $0.75 | $0.075 | $3.75 | Not listed |
Gemini 3.6 Flashgemini-3.6-flashFrom January 1, 2027: $1.50 in, $7.50 out. Free tier available. | $0.75 | $0.075 | $3.75 | Not listed |
Gemini 3.5 Flash-Litegemini-3.5-flash-liteFree tier available. | $0.30 | $0.03 | $2.50 | Not listed |
Gemini 3.1 Flash-Litegemini-3.1-flash-liteFree tier available. Text, image and video input. Audio input costs $0.50. | $0.25 | $0.025 | $1.50 | Not listed |
Gemini 3 Flash Previewgemini-3-flash-previewFree tier available. | $0.50 | $0.05 | $3.00 | Not listed |
Google Gemini pricing details, notes and price changes · Official source
DeepSeek API pricing
DeepSeek API pricing covers two current models. DeepSeek V4.1 Flash costs $0.30 per million input tokens and $1.20 per million output tokens at peak times, and V4 Pro costs $1.32 and $3.96. Off-peak rates are half price, and cache hits cost far less: $0.006 and $0.044 per million at peak. Both models have a 1M token context and up to 384K output tokens.
| Model | Input | Cached | Output | Context |
|---|---|---|---|---|
DeepSeek V4.1 Flashdeepseek-flashPeak price. Off-peak: $0.15 in, $0.60 out. Max output 384K tokens. | $0.30 | $0.006 | $1.20 | 1M |
DeepSeek V4 Prodeepseek-v4-proPeak price. Off-peak: $0.66 in, $1.98 out. Max output 384K tokens. | $1.32 | $0.044 | $3.96 | 1M |
DeepSeek pricing details, notes and price changes · Official source
Mistral API pricing
Mistral API pricing starts at $0.10 per million input and output tokens for Ministral 3 3B and goes up to $0.70 in and $2.10 out for Mistral Large 4, which has a 1M token context window. Mistral Small 4 costs $0.15 in and $0.60 out, and Codestral $0.30 in and $0.90 out. Cached input costs a tenth of the input price, and new accounts start in a free mode with included monthly usage.
| Model | Input | Cached | Output | Context |
|---|---|---|---|---|
Mistral Large 4mistral-large-4 | $0.70 | $0.07 | $2.10 | 1M |
Mistral Small 4mistral-small-2603 | $0.15 | $0.015 | $0.60 | 256K |
Codestralcodestral-2508 | $0.30 | $0.03 | $0.90 | 128K |
Ministral 3 14Bministral-14b-2512 | $0.20 | $0.02 | $0.20 | 256K |
Ministral 3 8Bministral-8b-2512 | $0.15 | $0.015 | $0.15 | 256K |
Ministral 3 3Bministral-3b-2512 | $0.10 | $0.01 | $0.10 | 256K |
Mistral pricing details, notes and price changes · Official source
FAQ
Which LLM API is the cheapest?
On our November 1, 2026 check, the cheapest models were Ministral 3 3B ($0.10 in, $0.10 out per 1M tokens), Ministral 3 8B ($0.15 in, $0.15 out per 1M tokens), Ministral 3 14B ($0.20 in, $0.20 out per 1M tokens). Google's Flash and Flash-Lite models also have a free tier.
How are LLM API prices calculated?
You pay per token, priced per million. Input tokens (your prompt, files and history) and output tokens (the answer, including any reasoning or thinking tokens) have separate prices, and output usually costs 5 to 8 times as much as input. Cached input, where the provider reuses a prompt prefix it has seen, is billed at a large discount.
How many tokens is a word?
For English text, a token is about 4 characters, or roughly three quarters of a word, so 1,000 words is about 1,300 tokens. Code and other languages often use more tokens. Each provider has its own tokenizer, so the same text can count differently.
How can I cut my LLM API costs?
Route easy tasks to a small model, keep a stable prompt prefix so it is cached, send work that can wait through a batch API (about half price at OpenAI, Anthropic, Google and Mistral), and limit output length. Long prompts can also cost more per token on some models.
How we check prices
We copy each price from the provider's official pricing page and note the date. "Not listed" means the page did not state it on that date; we never fill gaps from memory. Prices change often, so confirm on the source before you commit to a budget. Spotted a change? Tell us on the contact page. The same table is free to use as JSON; please link back to this page if you cite it.
- OpenAI API pricing
- OpenAI models
- Claude API pricing
- Claude models overview
- Claude pricing (claude.com)
- Gemini API pricing
- Gemini models
- DeepSeek API pricing
- DeepSeek API docs
- Mistral API pricing (docs)
- Mistral pricing
- Mistral models
- Mistral usage limits
Choosing a coding tool rather than an API? Read Claude Code vs Cursor.