DeepSeek V3.2 Thinking
| Released | 2025-12-01 |
|---|---|
| Price per 1M | $0.29 in · $0.43 out |
| Overall score | 51.0 · rank 252 |
| Context window | 128,000 tokens |
| Provider | DeepSeek |
| Weights | Open weights (MIT License) |
| Configuration | Thinking |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 51.0 | 252 | 5 |
| Overall: Coding | 50.8 | 249 | 2 |
| Overall: Reasoning | 50.0 | 265 | 2 |
| Overall: Agentic work | 53.5 | 228 | 1 |
| Overall: Frontend | 45.2 | 287 | 2 |
| Coding | 53.5 | 205 | 1 |
| Math | 49.5 | 186 | 1 |
| Knowledge | 50.8 | 120 | 1 |
| Frontend | 41.1 | 170 | 1 |
| General | 51.7 | 143 | 1 |
| Value vs price | +4.2 | 143 |
Evidence
Coding 53.5 ±5.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveCodeBench (Vals run) | 80.7% | Vals AI | 2026-09-01 |
Math 49.5 ±4.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MathArena Final-Answer Competitions | 57.7% | MathArena | 2026-11-11 |
Knowledge 50.8 ±4.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMLU-Pro (Vals run) | 84.9% | Vals AI | 2026-09-01 |
Frontend 41.1 ±3.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena WebDev | 1,363 | LMArena | 2026-11-09 |
General 51.7 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,423 | LMArena | 2026-11-09 |
Names matched to this model
deepseek-v3-2-thinking