C-Eval
C-Eval - comprehensive Chinese evaluation benchmark across 52 subjects.
Category
Knowledge
Max Score
100.0
Score Type
percent
Active
Yes
Model Scores
Scores for C-Eval
| Model | Score | Date | Verified |
|---|---|---|---|
| DeepSeek-V3.2 |
75.23%
|
26.06.2026 | Verified |
| DeepSeek-V4-Flash |
90.83%
|
26.06.2026 | Verified |
| DeepSeek-V4-Pro |
100.00%
|
26.06.2026 | Verified |
| Qwen3.6-35B-A3B |
71.56%
|
16.04.2026 | Verified |
| Gemma4-31B-it |
3.67%
|
16.04.2026 | Unverified |
| Qwen3.5-35B-A3B |
73.39%
|
16.04.2026 | Unverified |
| Gemma4-26B-A4B |
2.75%
|
16.04.2026 | Unverified |
| Qwen3.6-27B |
84.40%
|
22.04.2026 | Verified |
| Qwen3.5-397B-A17B |
99.08%
|
22.04.2026 | Unverified |
| Claude 4.5 Opus |
91.74%
|
22.04.2026 | Unverified |
| Qwen3.5-27B |
76.15%
|
16.04.2026 | Unverified |
| Qwen3.5-4B |
26.61%
|
01.02.2026 | Verified |
| Qwen3.5-122B-A10B |
88.99%
|
01.02.2026 | Verified |
| Qwen3-235B-A22B |
90.83%
|
01.02.2026 | Unverified |
| GPT-5-mini 2025-08-07 |
82.20
|
01.02.2026 | Unverified |