DeepSeek V3.2 vs GPT-5.2
Head to head on 29 shared benchmarks, with price, context window, and release dates.
Benchmark scores
47evals · sort by any column, ⤢ to expand fullscreen
| Model | ||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-5.2 OpenAI | 51.0% | 65.9% | 60.0% | 1699elo | 71.6 | 40.7% | 71.2% | 8.0% | 85.0% | 47.4% | 79.0% | 78.7% | 76.1% | 78.2% | 61.8% | 79.8% | 93.2% | 83.2% | 66.9% | 72.0% | 81.4% | 15.0% | 40.4% | 66.7% | 71.8% | 75.8% | 31.8% | 53.5% | 46.5% | - | - | 100 | 100.0% | 0.0% | 16.7% | - | 40.6% | 9.4% | 12.93 | 24.0% | 49.7% | 84.4% | 67.1% | 96.7 | 18.7% | 49.3% | - | 56.3% |
| DeepSeek V3.2 DeepSeek | 59.0% | 47.9% | 16.0% | 1511elo | 66.8 | 22.1% | 75.1% | 11.2% | 31.3% | 49.0% | 68.0% | 65.9% | 64.6% | 50.0% | 48.2% | 70.4% | 85.0% | 77.2% | 59.3% | 57.5% | 83.7% | 8.0% | 38.7% | 59.0% | 70.0% | 68.2% | 32.6% | 5.1% | 78.9% | 74.7% | 74.2% | - | - | - | - | 0.0% | - | - | - | - | - | - | - | - | - | - | 65 | 51.7% |
2 / 2 models
| Attribute | DeepSeek V3.2 DeepSeek | GPT-5.2 OpenAI |
|---|---|---|
| Description | DeepSeek V3.2 is an AI model from DeepSeek, released with open weights. | GPT-5.2 is an AI model from OpenAI. |
| Family | Deepseek | GPT |
| Params | - | - |
| Open weights | open | closed |
| License | mit | proprietary |
| Released | 1 Dec 2025 | 11 Dec 2025 |
| Context | 163,840 | 400,000 |
| Input $/1M | $0.28 | $1.75 |
| Output $/1M | $0.42 | $14.00 |
| Cache read $/1M | $0.130 | $0.175 |
| 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) |
FAQ
- Is DeepSeek V3.2 or GPT-5.2 better?
- Across the 29 benchmarks both models report on Sophon, GPT-5.2 leads on 22 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
- How do DeepSeek V3.2 and GPT-5.2 compare on GPQA Diamond?
- DeepSeek V3.2 scores 75.1% and GPT-5.2 scores 71.2% on GPQA Diamond.
- Which is cheaper, DeepSeek V3.2 or GPT-5.2?
- DeepSeek V3.2 is cheaper at $0.42 per million output tokens against $14.00 for GPT-5.2 - about 33.3x.
- When were DeepSeek V3.2 and GPT-5.2 released?
- DeepSeek V3.2 was released 1 Dec 2025 by DeepSeek. GPT-5.2 was released 11 Dec 2025 by OpenAI.
Comparing something else? Build your own side-by-side.