Model router
OpenRouter
A broad model catalog with observed input and output prices supplied by Wharf snapshots.
← All providers
566 models with current observed prices · 8 with published task results
| Model | Evidence | Cheapest passing | Median quality | Input / output per 1M |
|---|---|---|---|---|
| GPT-5.4 Frontier | 263 tasks | 69 | 93.0% | $2.5 / $15 |
| DeepSeek V3 Mid | 259 tasks | 12 | 66.7% | $0.29 / $1.14 |
| Qwen3 235B A22B Mid · 235B (22B active) | 259 tasks | 10 | 54.0% | $0.0875 / $0.35 |
| Granite 4.1 8B Small · 8B | 261 tasks | 6 | 38.0% | $0.05 / $0.1 |
| Claude Opus 4.7 Frontier | 259 tasks | 5 | 79.0% | $5 / $25 |
| Mistral Large 2407 Mid · 123B | 259 tasks | 5 | 66.0% | $2 / $6 |
| Arcee Trinity Large Thinking Mid · 400B | 259 tasks | 3 | 45.0% | $0.25 / $0.8 |
| Gemini 3.1 Pro Preview Frontier | 259 tasks | 0 | 54.0% | $2 / $12 |
| ~anthropic/claude-fable-latest Unclassified | Not yet measured | — | — | $10 / $50 |
| ~anthropic/claude-haiku-latest Unclassified | Not yet measured | — | — | $1 / $5 |
| ~anthropic/claude-opus-latest Unclassified | Not yet measured | — | — | $4 / $20 |
| ~anthropic/claude-sonnet-latest Unclassified | Not yet measured | — | — | $2 / $10 |
| ~deepseek/deepseek-flash-latest Unclassified | Not yet measured | — | — | $0.02 / $0.6 |
| ~deepseek/deepseek-pro-latest Unclassified | Not yet measured | — | — | $0.132 / $0.396 |
| ~deepseek/deepseek-v4-flash-latest Unclassified | Not yet measured | — | — | $0.0038 / $1.0436 |
| ~google/gemini-flash-latest Unclassified | Not yet measured | — | — | $0.75 / $3.75 |
| ~google/gemini-pro-latest Unclassified | Not yet measured | — | — | $2 / $12 |
| ~moonshotai/kimi-latest Unclassified | Not yet measured | — | — | $1 / $11.357 |
| ~openai/gpt-astra-latest Unclassified | Not yet measured | — | — | $10 / $50 |
| ~openai/gpt-latest Unclassified | Not yet measured | — | — | $2 / $10 |