DeepSeek V3.1
| Released | 2025-08-21 |
|---|---|
| Price per 1M | $0.25 in · $0.95 out |
| Overall score | 48.5 (early, few results) |
| Context window | 163,840 tokens |
| Provider | DeepSeek |
| Weights | Open weights |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 48.5 | 5 | |
| Overall: Coding | 46.7 | 2 | |
| Overall: Reasoning | 46.7 | 2 | |
| Overall: Agentic work | 46.7 | 2 | |
| Reasoning | 46.7 | 2 | |
| General | 51.2 | 3 | |
| Value vs price | +3.3 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 48.5 ±5.7 | 5 |
| Thinking | 50.7 ±7.0 | 1 |
Evidence
Reasoning 46.7 ±4.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleBench | 40.0% | SimpleBench | 2026-11-11 |
| Kagi Reasoning Benchmark | 53.2% | Kagi | 2026-11-11 |
General 51.2 ±3.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,417 | LMArena | 2026-11-09 |
| EQ-Bench Creative Writing v3 | 1,436 | EQ-Bench | 2026-11-11 |
| EQ-Bench Longform Creative Writing | 62.3% | EQ-Bench | 2026-11-11 |
Names matched to this model
deepseek-chat-v3-1