by DeepSeekRank #99Score 69.1
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
| Signal | Normalized | Weight | Contribution | Freshness |
|---|---|---|---|---|
Capabilities capability | 66.7 | 20% | 13.3 | 2026-06-24T09:59:29.571Z |
Benchmarks benchmark | 68.9 | 30% | 20.7 | 2026-06-24T09:59:29.571Z |
Pricing pricing_tier | 99.2 | 15% | 14.9 | 2026-06-24T09:59:29.571Z |
Context Window context_window | 74.5 | 10% | 7.4 | 2026-06-24T09:59:29.571Z |
Recency recency | 76.9 | 15% | 11.5 | 2026-06-24T09:59:29.571Z |
Output Capacity output_capacity | 75.3 | 10% | 7.5 | 2026-06-24T09:59:29.571Z |
| Capability | Supported |
|---|---|
| Vision | No |
| Reasoning | Yes |
| JSON Mode | Yes |
| Streaming | Yes |
| Function Calling | Yes |
| Web Search | No |