DeepSeek V3.1-Terminus vs DeepSeek V3.2
| Spec | DeepSeek V3.1-Terminus | DeepSeek V3.2 |
|---|---|---|
| Respan Index | 20.5 (#125) | 26.6 (#103) |
| Blended price per 1M | $0.453 | $0.29 |
| Context window | 128K | 131K |
| Input / output per 1M | $0.27 / $1 | $0.26 / $0.38 |
| Max output | - | 66K |
| Input | text | text |
| Released | 2025-09-22 | 2025-12-01 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | DeepSeek V3.1-Terminus | DeepSeek V3.2 |
|---|---|---|
| Coding | 4.7 DeepSeek V3.1-Terminus | 8.5 DeepSeek V3.2 |
| Agents | 14.8 DeepSeek V3.1-Terminus | 32.1 DeepSeek V3.2 |
| Reasoning | 25.8 DeepSeek V3.1-Terminus | 27.9 DeepSeek V3.2 |
| Knowledge & language | 19.7 DeepSeek V3.1-Terminus | 20.2 DeepSeek V3.2 |
| Human preference | 47.8 DeepSeek V3.1-Terminus | 50.0 DeepSeek V3.2 |
Benchmarks both report
DeepSeek V3.2 leads on 2 of the 3 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | DeepSeek V3.1-Terminus | DeepSeek V3.2 |
|---|---|---|
| LMArena Elo | 1416 | 1425 |
| Humanity's Last Exam (no tools) | 21.7% | 25.1% |
| MMLU-Pro | 85% | 85% |
FAQ
Is DeepSeek V3.1-Terminus better than DeepSeek V3.2?
DeepSeek V3.2 scores 26.6 on the Respan Index (#103), ahead of DeepSeek V3.1-Terminus at 20.5 (#125). DeepSeek V3.2 leads on 2 of the 3 benchmarks both report.
Which is cheaper, DeepSeek V3.1-Terminus or DeepSeek V3.2?
DeepSeek V3.1-Terminus lists at $0.27 input and $1 output per 1M tokens; DeepSeek V3.2 at $0.26 and $0.38. DeepSeek V3.2 is 1.6x cheaper per token at list price.
Which has the larger context window?
DeepSeek V3.1-Terminus accepts 128K tokens and DeepSeek V3.2 131K tokens.