← API pricing · cost calculator
LLM API cost calculator
Set the tokens in and out of one request and how many requests you send a day. Every one of the 39 live models with a published price is ranked by what that workload costs, cheapest first. The prices are each maker's list price on its own API, from the models.dev catalog as of October 25, 2026; the full table is on LLM API pricing.
Cheapest: MiMo-V2.5 at $9.66 a month · Top of the bench index: Claude Opus 5.5 at $420 (30,000 requests over 30 days)
| # | Model | Per request | Per 1K requests | Per month | AW Index |
|---|---|---|---|---|---|
| 1 | MiMo-V2.5Xiaomi · $0.14 / $0.28 per 1M | $0.00032 | $0.32 | $9.66 | — |
| 2 | Claude Haiku 5.5Anthropic · $0.10 / $0.50 per 1M | $0.00035 | $0.35 | $10.50 | 50.5 |
| 3 | GPT-6 LunaOpenAI · $0.10 / $0.50 per 1M | $0.00035 | $0.35 | $10.50 | 28 |
| 4 | Qwen 3.8 FlashAlibaba · $0.15 / $0.47 per 1M | $0.00041 | $0.41 | $12.39 | — |
| 5 | GLM-5.3 FlashZhipu AI · $0.15 / $0.50 per 1M | $0.00042 | $0.42 | $12.75 | 32.4 |
| 6 | MiniMax M3.1 FlashMiniMax · $0.15 / $0.60 per 1M | $0.00047 | $0.47 | $13.95 | 48.8 |
| 7 | MiniMax M3MiniMax · $0.30 / $1.20 per 1M | $0.00093 | $0.93 | $27.90 | 5.8 |
| 8 | MiMo-V2.6 ProXiaomi · $0.435 / $0.87 per 1M | $0.001 | $1.00 | $30.02 | 65.1 |
| 9 | Gemini 3.5 Flash-LiteGoogle · $0.30 / $2.50 per 1M | $0.0014 | $1.45 | $43.50 | — |
| 10 | DeepSeek V4 ProDeepSeek · $0.66 / $1.98 per 1M | $0.0018 | $1.78 | $53.46 | 31.8 |
| 11 | Gemini 3.6 FlashGoogle · $0.75 / $3.75 per 1M | $0.0026 | $2.63 | $78.75 | — |
| 12 | Gemini 3.7 FlashGoogle · $0.75 / $3.75 per 1M | $0.0026 | $2.63 | $78.75 | — |
| 13 | Gemini 3.8 FlashGoogle · $0.75 / $3.75 per 1M | $0.0026 | $2.63 | $78.75 | 43.3 |
| 14 | Muse Spark 1.1Meta · $1.25 / $4.25 per 1M | $0.0036 | $3.58 | $107 | 30.8 |
| 15 | Muse Spark 1.2Meta · $1.25 / $4.25 per 1M | $0.0036 | $3.58 | $107 | — |
| 16 | Muse Spark 1.3Meta · $1.25 / $4.25 per 1M | $0.0036 | $3.58 | $107 | 74.9 |
| 17 | Mistral Large 4Mistral · $1.36 / $4.18 per 1M | $0.0037 | $3.71 | $111 | 32.6 |
| 18 | GLM-5.2Zhipu AI · $1.40 / $4.40 per 1M | $0.0039 | $3.86 | $116 | 21.2 |
| 19 | GLM-5.3Zhipu AI · $1.40 / $4.40 per 1M | $0.0039 | $3.86 | $116 | 60 |
| 20 | Grok 4.5xAI · $2 / $6 per 1M | $0.0054 | $5.40 | $162 | 34.9 |
| 21 | Grok 4.6xAI · $2 / $6 per 1M | $0.0054 | $5.40 | $162 | 58.1 |
| 22 | Grok 4.7xAI · $2 / $6 per 1M | $0.0054 | $5.40 | $162 | 59.4 |
| 23 | Qwen 3.8 MaxAlibaba · $2 / $6 per 1M | $0.0054 | $5.40 | $162 | 62.9 |
| 24 | Gemini 3.5 FlashGoogle · $1.50 / $9 per 1M | $0.0059 | $5.85 | $176 | 21.3 |
| 25 | Claude Sonnet 5Anthropic · $2 / $10 per 1M | $0.007 | $7.00 | $210 | 27.9 |
| 26 | Claude Sonnet 5.5Anthropic · $2 / $10 per 1M | $0.007 | $7.00 | $210 | 93.5 |
| 27 | Gemini 4 ArgonGoogle · $2 / $10 per 1M | $0.007 | $7.00 | $210 | 79.5 |
| 28 | GPT-6 SolOpenAI · $2 / $10 per 1M | $0.007 | $7.00 | $210 | 68.6 |
| 29 | GPT-6.1 SolOpenAI · $2 / $10 per 1M | $0.007 | $7.00 | $210 | 82.9 |
| 30 | Kimi K3.1 ProMoonshot · $2 / $10 per 1M | $0.007 | $7.00 | $210 | 86 |
| 31 | Kimi K3Moonshot · $3 / $15 per 1M | $0.01 | $10.50 | $315 | 27.3 |
| 32 | ChatGPT-5.6OpenAI · $4 / $20 per 1M | $0.01 | $14.00 | $420 | 50 |
| 33 | Claude Opus 5.5Anthropic · $4 / $20 per 1M | $0.01 | $14.00 | $420 | 100 |
| 34 | Claude Opus 4.8Anthropic · $5 / $25 per 1M | $0.02 | $17.50 | $525 | 37.2 |
| 35 | Claude Opus 5Anthropic · $5 / $25 per 1M | $0.02 | $17.50 | $525 | 72.8 |
| 36 | GPT-5.5OpenAI · $5 / $30 per 1M | $0.02 | $19.50 | $585 | 42.6 |
| 37 | Fable 5Anthropic · $10 / $50 per 1M | $0.04 | $35.00 | $1,050 | 72.1 |
| 38 | Fable 5.1Anthropic · $10 / $50 per 1M | $0.04 | $35.00 | $1,050 | 90.7 |
| 39 | GPT-6 AstraOpenAI · $10 / $50 per 1M | $0.04 | $35.00 | $1,050 | 87.4 |
Quick answers
How much does an LLM API cost for 1,000 chat requests a day?
At 1,500 input and 400 output tokens each, that is 30,000 requests over 30 days. The cheapest of the 39 priced models is MiMo-V2.5 at $9.66 a month, and the most expensive is GPT-6 Astra at $1,050. Claude Opus 5.5, the top model on the ArtificialWatch bench index, costs $420. List prices as of October 25, 2026.
How do you calculate LLM API cost?
Per request: input tokens times the input price plus output tokens times the output price, each price per million tokens, divided by one million. Multiply by requests for a total. Input read from a prompt cache is billed at the cached-input price where the maker publishes one, and a prompt past a model's long-prompt threshold is billed at its long-prompt rates for the whole request. Most reasoning models bill their thinking tokens as output, so count them in the output figure.
Are these the prices I will pay?
They are each maker's published list price on its own API, from the models.dev catalog as of October 25, 2026. Batch discounts, committed-use deals, free tiers, taxes and reseller or cloud prices are not included, so treat a result as an estimate, not a quote.
Cheap is one axis. For how good each model is, see the bench index and the coding leaderboard; for what shipped when, new AI model releases. Prices change when a new model lands. To hear the minute one goes live, watch free.
Prices move. Hear it the minute a new model lands.
Free: a Chrome push and an email about 15 minutes after a new model answers on a public API. Members get a text or a phone call the moment a launch is confirmed.
- Free forever
- No card
- Unsubscribe in one click