Estimate your monthly bill for Claude, GPT, Gemini, DeepSeek and 40+ AI models from your real token usage. Free and 100% client-side.
| Model | Vendor | Price in/out ($/M) | Monthly (USD) |
|---|---|---|---|
| Cursor Agent | Cursor | $0 / $0 | $0.00 |
| GitHub Copilot | GitHub / Microsoft | $0 / $0 | $0.00 |
| Llama 4 Scout | Meta | $0.18 / $0.6 | $5.70 |
| Mistral Small 3 | Mistral AI | $0.2 / $0.6 | $6.00 |
| Codestral | Mistral AI | $0.3 / $0.9 | $9.00 |
| DeepSeek V3 | DeepSeek | $0.27 / $1.1 | $9.55 |
| Llama 4 Maverick | Meta | $0.35 / $1.4 | $12.25 |
| GPT-5 mini | OpenAI | $0.25 / $2 | $13.75 |
| Qwen 3 | Alibaba | $0.4 / $1.6 | $14.00 |
| Gemini 3 Flash | Google DeepMind | $0.3 / $2.5 | $17.00 |
| Kimi K2 | Moonshot AI | $0.5 / $2 | $17.50 |
| DeepSeek R1 | DeepSeek | $0.55 / $2.19 | $19.20 |
Benchmarks and rankings for every model in the AI comparison
Every price is the official list price per million tokens published by each provider (Anthropic, OpenAI, Google, DeepSeek, Mistral and others). The dataset is the same one that powers our AI model comparison hub and is refreshed on every deploy.
Monthly cost = requests per month x (input tokens x input price + output tokens x output price) / 1,000,000. It assumes list prices without volume discounts, batch APIs or cached-prompt savings, so treat it as a realistic upper bound.
Yes. Prompt caching can cut input costs by up to 90% for repeated system prompts, batch APIs typically halve the price for non-urgent workloads, and routing simple requests to a smaller model of the same family is the single biggest saving in production systems.
No. The calculator runs entirely in your browser. Nothing you type is stored or transmitted.