DeepSeek V3.2 vs GLM 5
Head to head on 23 shared benchmarks, with price, context window, and release dates.
Benchmark scores
39evals · sort by any column, ⤢ to expand fullscreen
| Model | ||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GLM 5 Zai | 62.9% | 1598elo | 70.9 | 16.4% | 66.6% | 7.6% | 55.2% | 82.7% | 71.2% | 73.6% | 67.9% | 55.3% | 77.5% | 83.5% | 69.1% | 65.2% | 38.3% | 69.7% | 72.8% | 70.0% | 39.4% | 23.4% | 97.4% | - | - | - | - | - | 7.9% | 24.7% | 97.9 | - | 48395 | - | - | - | 73.3 | 61.12 | - | 56.4% |
| DeepSeek V3.2 DeepSeek | 47.9% | 1511elo | 66.8 | 22.1% | 75.1% | 11.2% | 49.0% | 68.0% | 65.9% | 64.6% | 50.0% | 48.2% | 70.4% | 85.0% | 77.2% | 57.5% | 38.7% | 59.0% | 70.0% | 68.2% | 32.6% | 5.1% | 78.9% | 74.7% | 74.2% | 59.0% | 0.0% | 16.0% | - | - | - | 31.3% | - | 59.3% | 83.7% | 8.0% | - | - | 65 | 51.7% |
2 / 2 models
| Attribute | DeepSeek V3.2 DeepSeek | GLM 5 Zai |
|---|---|---|
| Description | DeepSeek V3.2 is an AI model from DeepSeek, released with open weights. | GLM-5 is an AI model from Zai, released with open weights. |
| Family | Deepseek | Glm |
| Params | - | - |
| Open weights | open | open |
| License | mit | mit |
| Released | 1 Dec 2025 | 11 Feb 2026 |
| Context | 163,840 | 204,800 |
| Input $/1M | $0.28 | $1.00 |
| Output $/1M | $0.42 | $3.20 |
| Cache read $/1M | $0.130 | $0.120 |
| Cache write $/1M | - | - |
| Speed | 0 tok/s | 0 tok/s |
| Latency | 0.00 s | 0.00 s |
| Modalities | text | text |
| Providers | - | - |
| Publisher | DeepSeek | Zai |
| Reported scores | Aider Polyglot Benchmark74.2% (Aider) MathArena57.47% GPQA Diamond75.1% (AA) Humanity's Last Exam (HLE)11.2% (AA) | GPQA Diamond66.6% (AA) Humanity's Last Exam (HLE)7.6% (AA) IFBench55.2% (AA) SciCode38.3% (AA) τ²-bench (Tau²-bench)97.4% (AA) |
FAQ
- Is DeepSeek V3.2 or GLM 5 better?
- Across the 23 benchmarks both models report on Sophon, GLM 5 leads on 17 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
- How do DeepSeek V3.2 and GLM 5 compare on GPQA Diamond?
- DeepSeek V3.2 scores 75.1% and GLM 5 scores 66.6% on GPQA Diamond.
- Which is cheaper, DeepSeek V3.2 or GLM 5?
- DeepSeek V3.2 is cheaper at $0.42 per million output tokens against $3.20 for GLM 5 - about 7.6x.
- When were DeepSeek V3.2 and GLM 5 released?
- DeepSeek V3.2 was released 1 Dec 2025 by DeepSeek. GLM 5 was released 11 Feb 2026 by Zai.
Comparing something else? Build your own side-by-side.