Claude Sonnet 4.5 vs DeepSeek V3.2
Head to head on 27 shared benchmarks, with price, context window, and release dates.
Benchmark scores
88evals · sort by any column, ⤢ to expand fullscreen
| Model | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V3.2 DeepSeek | 92.0% | 47.9% | 1515elo | 66.8 | 22.1% | 84.0% | 24.6% | 31.3% | 60.7% | 68.0% | 65.9% | 64.6% | 50.0% | 48.2% | 70.4% | 85.0% | 77.2% | 86.2% | 86.2% | 8.0% | 38.9% | 59.0% | 70.0% | 68.2% | 35.6% | 5.1% | 90.6% | - | 74.7% | - | 74.2% | - | - | - | - | - | - | 0.0% | - | - | 13.0% | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | 57.5% | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | 65 | - | - | - | 55.3% |
| Claude Sonnet 4.5 Anthropic | 88.0% | 60.8% | 1678elo | 71.4 | 15.2% | 83.4% | 17.8% | 87.5% | 57.3% | 83.8% | 70.7% | 80.4% | 57.0% | 53.4% | 76.5% | 79.3% | 77.6% | 71.4% | 87.5% | 19.0% | 45.7% | 67.0% | 74.8% | 73.3% | 35.6% | 22.6% | 78.1% | 0.0% | - | 0.0% | - | 100 | 0.0% | 0.0% | 0.0% | 5.0% | 0.0% | - | 0.0% | 0.0% | - | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 83.9% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.75 | 0.75 | 0.0% | - | 0 | 40.6% | 84.5% | 0 | 64.0% | 0.0% | 0.0% | 62.9% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 32.9% | 0.0% | 0 | 0.0% | 0.0% | 77.2% | 0.0% | 0 | 0 | 50.34 | - | 0 | 0.0% | 1.49 | 29.2% |
2 / 2 models
| Attribute | Claude Sonnet 4.5 Anthropic | DeepSeek V3.2 DeepSeek |
|---|---|---|
| Description | anthropic/claude-sonnet-4.5 is an AI model. | DeepSeek V3.2 is an AI model from DeepSeek, released with open weights. |
| Family | Claude | Deepseek |
| Params | Undisclosed (closed) | - |
| Open weights | closed | open |
| License | proprietary | mit |
| Released | 29 Sep 2025 | 1 Dec 2025 |
| Context | 1,000,000 | 163,840 |
| Input $/1M | $3.00 | $0.28 |
| Output $/1M | $15.00 | $0.42 |
| Cache read $/1M | $0.300 | $0.140 |
| Cache write $/1M | $3.750 | - |
| Speed | 0 tok/s | 0 tok/s |
| Latency | 0.00 s | 0.00 s |
| Modalities | text, image, file | text |
| Providers | anthropic, aws-bedrock, vertex-ai | - |
| Publisher | Anthropic | DeepSeek |
| Reported scores | TaxEval v273.3% LiveBench - Math79.3% (LiveBench) LiveBench - Coding80.4% (LiveBench) LiveBench - Language76.5% (LiveBench) LiveBench - Data Analysis57.0% (LiveBench) | Aider Polyglot Benchmark74.2% (Aider) MathArena57.47% GPQA Diamond84.0% (AA) Humanity's Last Exam (HLE)24.6% (AA) |
FAQ
- Is Claude Sonnet 4.5 or DeepSeek V3.2 better?
- Across the 27 benchmarks both models report on Sophon, Claude Sonnet 4.5 leads on 18 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
- How do Claude Sonnet 4.5 and DeepSeek V3.2 compare on GPQA Diamond?
- Claude Sonnet 4.5 scores 83.4% and DeepSeek V3.2 scores 84.0% on GPQA Diamond.
- Which is cheaper, Claude Sonnet 4.5 or DeepSeek V3.2?
- DeepSeek V3.2 is cheaper at $0.42 per million output tokens against $15.00 for Claude Sonnet 4.5 - about 35.7x.
- When were Claude Sonnet 4.5 and DeepSeek V3.2 released?
- Claude Sonnet 4.5 was released 29 Sep 2025 by Anthropic. DeepSeek V3.2 was released 1 Dec 2025 by DeepSeek.
Other head-to-heads
56Claude Sonnet 4.5 vs GPT-4.1 Mini35 shared evalsClaude Opus 4.5 vs Claude Sonnet 4.534 shared evalsClaude Sonnet 4.5 vs GPT-5.232 shared evalsClaude Sonnet 4.5 vs GPT-5 Mini31 shared evalsClaude Sonnet 4.5 vs GPT-531 shared evalsClaude Sonnet 4.5 vs GPT-5 Nano31 shared evalsDeepSeek V3.2 vs GPT-5.229 shared evalsClaude 4 Sonnet vs Claude Sonnet 4.528 shared evalsClaude Opus 4.6 vs Claude Sonnet 4.528 shared evalsClaude Sonnet 4.5 vs Kimi K2.528 shared evalsClaude Sonnet 4.5 vs Grok 4.1 Fast28 shared evalsClaude Sonnet 4.5 vs Gemini 3.1 Pro Preview27 shared evalsClaude Opus 4.5 vs DeepSeek V3.227 shared evalsClaude Sonnet 4.5 vs GPT-5.526 shared evalsClaude Sonnet 4.5 vs GPT-5.426 shared evalsClaude Sonnet 4.5 vs Gemini 2.5 Pro26 shared evalsDeepSeek V3.2 vs GPT-5 Mini26 shared evalsClaude Opus 4.6 vs DeepSeek V3.226 shared evalsDeepSeek V3.2 vs Grok 4.1 Fast26 shared evalsClaude Opus 4.8 vs Claude Sonnet 4.525 shared evalsClaude Sonnet 4.5 vs GLM 4.725 shared evalsDeepSeek V3.2 vs Kimi K2.525 shared evalsClaude Fable 5 vs Claude Sonnet 4.524 shared evalsClaude Opus 4.7 vs Claude Sonnet 4.524 shared evalsClaude Sonnet 4.5 vs Claude Sonnet 4.624 shared evalsDeepSeek V3.2 vs Gemini 3.1 Pro Preview24 shared evalsDeepSeek V3.2 vs GPT-5.424 shared evalsDeepSeek V3.2 vs GLM 5.124 shared evalsClaude Sonnet 4.5 vs GLM 5.123 shared evalsClaude Sonnet 4.5 vs gpt-oss-120b23 shared evalsClaude Sonnet 4.5 vs GPT-5.4 Mini23 shared evalsDeepSeek V3.2 vs GPT-5.523 shared evalsClaude Opus 4.7 vs DeepSeek V3.223 shared evalsClaude Sonnet 4.6 vs DeepSeek V3.223 shared evalsDeepSeek V3.2 vs gpt-oss-120b23 shared evalsClaude Sonnet 4.5 vs Qwen3.6 Plus22 shared evalsDeepSeek V3.2 vs GPT-5 Nano22 shared evalsClaude Opus 4.8 vs DeepSeek V3.222 shared evalsDeepSeek V3.2 vs GPT-5.4 Mini22 shared evalsDeepSeek V3.2 vs GLM 4.722 shared evalsClaude Sonnet 4.5 vs GPT-4o21 shared evalsClaude 4 Sonnet vs DeepSeek V3.221 shared evalsDeepSeek V3.2 vs Gemini 2.5 Pro21 shared evalsClaude Fable 5 vs DeepSeek V3.220 shared evalsDeepSeek V3.2 vs Qwen3.6 Plus20 shared evalsClaude Sonnet 4.5 vs GPT-4.119 shared evalsDeepSeek V3.2 vs GPT-519 shared evalsDeepSeek V3.2 vs GPT-4o18 shared evalsClaude Sonnet 4.5 vs GPT-4o-mini17 shared evalsClaude Sonnet 4.5 vs GPT-4.1 Nano16 shared evalsDeepSeek V3.2 vs GPT-4.1 Mini15 shared evalsDeepSeek V3.2 vs GPT-4.115 shared evalsDeepSeek V3.2 vs GPT-4o-mini15 shared evalsDeepSeek V3.2 vs GPT-4.1 Nano15 shared evalsClaude Sonnet 4.5 vs GPT-6 Astra11 shared evalsDeepSeek V3.2 vs GPT-6 Astra9 shared evals
Comparing something else? Build your own side-by-side.