Compare models
Pick up to 4. Every model here is free in free-pi.
| DeepSeek V4 Pro | Claude Haiku 5.5 | Defense Flash | Kimi K3.1 Pro | |
|---|---|---|---|---|
| Maker | DeepSeek | Anthropic | — | MoonshotAI |
| Usage in free-pi | ~8.6x usage | ~2.0x usage | ~20x usage | ~41x usage |
| Intelligence index | 30.4 | 43.4 | — | 55.1 (best) |
| Coding index | 59.4 | — | — | — |
| Context window | 1,048,576 tokens (best) | 1,000,000 tokens | — | 1,048,576 tokens (best) |
| Max output | 384,000 tokens | 128,000 tokens | — | 943,718 tokens (best) |
| Inputs | Text | Text, image, file (best) | — | Text, image, video (best) |
| Tool calling | Yes | Yes | — | Yes |
| Reasoning effort | Yes | Yes | — | Yes |
| Structured output | Yes | Yes | — | Yes |
| Released | Apr 24, 2026 | Oct 7, 2026 | — | Oct 23, 2026 |
| Model page | Model page | Model page | Model page | Model page |
Stats from OpenRouter; refreshed automatically. Usage rate from free-pi. Green marks the best value in each row. “—” means no figure is available.