DeepSeek V3.2 vs GPT-5.5
Head to head on 23 shared benchmarks, with price, context window, and release dates.
Benchmark scores
62evals · sort by any column, ⤢ to expand fullscreen
| Model | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-5.5 OpenAI | 68.4% | 86.0% | 1844elo | 78.2 | 51.7% | 76.8% | 13.7% | 46.1% | 84.7% | 82.5% | 81.1% | 73.0% | 87.7% | 96.3% | 87.7% | 92.8% | 88.6% | 50.0% | 47.3% | 75.0% | 49.2% | 69.8% | 69.3% | - | - | - | 86.2% | 54.2% | 12.9% | 36.2% | - | 90.0% | 100.0% | 1315elo | 34.0% | 51.8% | 5.7% | 44.8% | 24.9% | 1769elo | 15.1% | 51.8% | - | - | 2.1% | - | 49.1% | 86.9% | 68.8% | 78.7% | 45.3% | 100.0% | 51.5% | 4.2 | - | - | 58.6% | 75.7% | 83.4% | - | 57.4% | 68.1% | 61.9% |
| DeepSeek V3.2 DeepSeek | 47.9% | 16.0% | 1511elo | 66.8 | 22.1% | 75.1% | 11.2% | 49.0% | 65.9% | 64.6% | 50.0% | 48.2% | 70.4% | 85.0% | 77.2% | 57.5% | 83.7% | 8.0% | 38.7% | 68.2% | 32.6% | 5.1% | 78.9% | 74.7% | 74.2% | 59.0% | - | - | - | - | 0.0% | - | - | - | - | - | - | - | - | - | - | - | 31.3% | 68.0% | - | 59.3% | - | - | - | - | - | - | - | - | 59.0% | 70.0% | - | - | - | 65 | - | - | 51.7% |
2 / 2 models
| Attribute | DeepSeek V3.2 DeepSeek | GPT-5.5 OpenAI |
|---|---|---|
| Description | DeepSeek V3.2 is an AI model from DeepSeek, released with open weights. | GPT-5.5 is an AI model from OpenAI. |
| Family | Deepseek | GPT |
| Params | - | - |
| Open weights | open | closed |
| License | mit | proprietary |
| Released | 1 Dec 2025 | 23 Apr 2026 |
| Context | 163,840 | 1,050,000 |
| Input $/1M | $0.28 | $5.00 |
| Output $/1M | $0.42 | $30.00 |
| Cache read $/1M | $0.130 | $0.500 |
| Cache write $/1M | - | - |
| Speed | 0 tok/s | 0 tok/s |
| Latency | 0.00 s | 0.00 s |
| Modalities | text | file, image, text |
| Providers | - | - |
| Publisher | DeepSeek | OpenAI |
| Reported scores | Aider Polyglot Benchmark74.2% (Aider) MathArena57.47% GPQA Diamond75.1% (AA) Humanity's Last Exam (HLE)11.2% (AA) | GPQA Diamond76.8% (AA) Humanity's Last Exam (HLE)13.7% (AA) IFBench46.1% (AA) SciCode47.3% (AA) τ²-bench (Tau²-bench)69.3% (AA) |
FAQ
- Is DeepSeek V3.2 or GPT-5.5 better?
- Across the 23 benchmarks both models report on Sophon, GPT-5.5 leads on 21 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
- How do DeepSeek V3.2 and GPT-5.5 compare on GPQA Diamond?
- DeepSeek V3.2 scores 75.1% and GPT-5.5 scores 76.8% on GPQA Diamond.
- Which is cheaper, DeepSeek V3.2 or GPT-5.5?
- DeepSeek V3.2 is cheaper at $0.42 per million output tokens against $30.00 for GPT-5.5 - about 71.4x.
- When were DeepSeek V3.2 and GPT-5.5 released?
- DeepSeek V3.2 was released 1 Dec 2025 by DeepSeek. GPT-5.5 was released 23 Apr 2026 by OpenAI.
Comparing something else? Build your own side-by-side.