Claude Sonnet 4 vs DeepSeek V3.2-Exp
| Spec | Claude Sonnet 4 | DeepSeek V3.2-Exp |
|---|---|---|
| Respan Index | 21.0 (#122) | 21.8 (#120) |
| Blended price per 1M | - | $0.305 |
| Context window | 200K | 128K |
| Input / output per 1M | - | $0.27 / $0.41 |
| Max output | 64K | - |
| Input | text, vision | text |
| Released | 2025-05-22 | 2025-09-29 |
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 | Claude Sonnet 4 | DeepSeek V3.2-Exp |
|---|---|---|
| Coding | 8.3 Claude Sonnet 4 | 4.9 DeepSeek V3.2-Exp |
| Agents | 17.3 Claude Sonnet 4 | 21.4 DeepSeek V3.2-Exp |
| Reasoning | 23.7 Claude Sonnet 4 | 24.4 DeepSeek V3.2-Exp |
| Knowledge & language | 17.7* Claude Sonnet 4 | 19.8 DeepSeek V3.2-Exp |
| Human preference | 46.1 Claude Sonnet 4 | 47.6 DeepSeek V3.2-Exp |
* Estimated: no published scores in that category. How the Respan Index works
Benchmarks both report
They split the 2 benchmarks both report evenly. Higher is better. Each score links to where it was published.
| Benchmark | Claude Sonnet 4 | DeepSeek V3.2-Exp |
|---|---|---|
| LMArena Elo | 1402 | 1425 |
| WeirdML (Håvard Tveit Ihle) | 46.1% | 39.5% |
FAQ
Is Claude Sonnet 4 better than DeepSeek V3.2-Exp?
DeepSeek V3.2-Exp scores 21.8 on the Respan Index (#120), ahead of Claude Sonnet 4 at 21.0 (#122). They split the 2 benchmarks both report evenly.
Which has the larger context window?
Claude Sonnet 4 accepts 200K tokens and DeepSeek V3.2-Exp 128K tokens.