LLM API pricing by model (August 2026)
Ranked by cost per task on a fixed profile: 1,500 input / 500 output tokens per call, 3 calls per task. Computed from the open dataset.
| # | Model | Provider | Tier | Input $/1M | Output $/1M | Output share | $/task | $/1M tasks |
|---|---|---|---|---|---|---|---|---|
| 1 | Mistral Small | Mistral | value | $0.15 | $0.6 | 57% | $0.00158 | $1,575 |
| 2 | DeepSeek-V4-Flash | DeepSeek | value | $0.44 | $1.32 | 50% | $0.00396 | $3,960 |
| 3 | Gemini 2.5 Flash | fast | $0.3 | $2.5 | 74% | $0.00510 | $5,100 | |
| 4 | Grok 4.3 | xAI | mid | $1.25 | $2.5 | 40% | $0.00938 | $9,375 |
| 5 | GPT-5.4 mini | OpenAI | mid | $0.75 | $4.5 | 67% | $0.01013 | $10,125 |
| 6 | Claude Haiku 4.5 | Anthropic | fast | $1 | $5 | 63% | $0.01200 | $12,000 |
| 7 | Claude Sonnet 5 | Anthropic | mid | $2 | $10 | 63% | $0.02400 | $24,000 |
| 8 | Gemini 3.1 Pro | frontier | $2 | $12 | 67% | $0.02700 | $27,000 | |
| 9 | Claude Opus 4.8 | Anthropic | frontier | $5 | $25 | 63% | $0.06000 | $60,000 |
| 10 | GPT-5.5 | OpenAI | frontier | $5 | $30 | 67% | $0.06750 | $67,500 |
The spread top to bottom is 42.9x for identical work. Before optimising anything else, check whether a cheaper model passes your evaluation cases: the cost-cutting guide ranks every lever by measured return, and the head-to-head pages put exact numbers on any two models.