Leaderboard
C-Eval
cevalRecord
- Category
- knowledge
- Scored models
- 64
Scores
| Model | Score |
|---|---|
| deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| deepseek-ai/DeepSeek-V4-Pro | 90.4 |
| deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| — | 90.4 |
| deepseek-ai/DeepSeek-V4-Pro | 90.4 |
| inclusionAI/LLaDA2.0-flash | 88.17 |
| inclusionAI/LLaDA2.0-flash-preview | 88.03 |
| Qwen/Qwen2-57B-A14B | 83.5 |
| SUSTech/SUS-Chat-34B | 82.42 |
| inclusionAI/LLaDA2.0-mini-preview | 82.1 |
| 01-ai/Yi-34B-200K | 81.9 |
| larryvrh/Yi-34B-200K-Llamafied | 81.9 |
| Qwen/Qwen2-72B | 81.7 |
| inclusionAI/LLaDA2.0-mini | 81.36 |
| Qwen/Qwen-72B | 80.1 |
| Qwen/Qwen-72B | 80.1 |
| Qwen/Qwen-72B | 80.1 |
| zai-org/glm-4-9b | 77.1 |
| zai-org/glm-4-9b-hf | 77.1 |
| 01-ai/Yi-34B-Chat | 77.04 |
| 01-ai/Yi-34B-Chat | 77.04 |
| Qwen/Qwen2-57B-A14B | 76.7 |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- category
- knowledge
- id
- ceval
- name
- C-Eval
- schema version
- local-ai-registry/v1
rows
| variant | conf | context | model id | org | rank | root | score |
|---|---|---|---|---|---|---|---|
| DeepSeek-V4-Flash | medium | C-Eval (EM) | deepseek-v4-flash | deepseek-ai | 1 | deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| DeepSeek-V4-Pro | medium | C-Eval (EM) | deepseek-v4-pro | deepseek-ai | 2 | deepseek-ai/DeepSeek-V4-Pro | 90.4 |
| DeepSeek-V4-Flash-DSpark | medium | C-Eval (EM) | deepseek-v4-flash | deepseek-ai | 3 | deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| DeepSeek-V4-Flash-GGUF | medium | C-Eval (EM) | deepseek-v4-flash | unsloth | 4 | deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| DeepSeek-V4-Flash | medium | C-Eval (EM) | deepseek-v4-flash | unsloth | 5 | deepseek-ai/DeepSeek-V4-Flash | 90.4 |
| DeepSeek-V4-Pro-DSpark | medium | C-Eval (EM) | Unknown | deepseek-ai | 6 | — | 90.4 |
| DeepSeek-V4-Pro | medium | C-Eval (EM) | deepseek-v4-pro | unsloth | 7 | deepseek-ai/DeepSeek-V4-Pro | 90.4 |
| LLaDA2.0-flash | medium | C-EVAL | llada2-0-flash | inclusionAI | 8 | inclusionAI/LLaDA2.0-flash | 88.17 |
| LLaDA2.0-flash-preview | medium | C-EVAL | llada2-0-flash-preview | inclusionAI | 9 | inclusionAI/LLaDA2.0-flash-preview | 88.03 |
| Qwen2-57B-A14B | medium | C-Eval | qwen2-57b-a14b | Qwen | 10 | Qwen/Qwen2-57B-A14B | 83.5 |
| SUS-Chat-34B-GGUF | high | ceval-chat | Unknown | TheBloke | 11 | SUSTech/SUS-Chat-34B | 82.42 |
| LLaDA2.0-mini-preview | medium | C-EVAL | llada2-0-mini-preview | inclusionAI | 12 | inclusionAI/LLaDA2.0-mini-preview | 82.1 |
| Yi-34B-200K-GGUF | high | C-Eval | yi-34b-200k | TheBloke | 13 | 01-ai/Yi-34B-200K | 81.9 |
| Yi-34B-200K-Llamafied-GGUF | high | C-Eval | yi-34b-200k-llamafied | TheBloke | 14 | larryvrh/Yi-34B-200K-Llamafied | 81.9 |
| Qwen2-72B | medium | C-Eval | qwen2-72b | Qwen | 15 | Qwen/Qwen2-72B | 81.7 |
| LLaDA2.0-mini | medium | C-EVAL | llada2-0-mini | inclusionAI | 16 | inclusionAI/LLaDA2.0-mini | 81.36 |
| Qwen-72B-Chat | low | CEval (val) | qwen-72b | Qwen | 17 | Qwen/Qwen-72B | 80.1 |
| Qwen-72B-Chat-Int4 | high | CEval (val) | qwen-72b | Qwen | 18 | Qwen/Qwen-72B | 80.1 |
| Qwen-72B-Chat-Int8 | high | CEval (val) | qwen-72b | Qwen | 19 | Qwen/Qwen-72B | 80.1 |
| glm-4-9b | high | C-Eval | glm-4-9b | zai-org | 20 | zai-org/glm-4-9b | 77.1 |
| glm-4-9b-hf | high | C-Eval | glm-4-9b-hf | zai-org | 21 | zai-org/glm-4-9b-hf | 77.1 |
| Yi-34B-Chat-GGUF | high | C-Eval(val)<sup>*</sup> | yi-34b-chat | TheBloke | 22 | 01-ai/Yi-34B-Chat | 77.04 |
| Yi-34B-Chat-AWQ | high | C-Eval(val)<sup>*</sup> | yi-34b-chat | TheBloke | 23 | 01-ai/Yi-34B-Chat | 77.04 |
| Qwen2-57B-A14B-Instruct | medium | C-Eval | qwen2-57b-a14b | Qwen | 24 | Qwen/Qwen2-57B-A14B | 76.7 |
| glm-4-9b-chat | high | C-Eval | glm-4-9b-chat | zai-org | 25 | zai-org/glm-4-9b-chat | 75.6 |
| glm-4-9b-chat-hf | high | C-Eval | glm-4-9b-chat | zai-org | 26 | THUDM/glm-4-9b-chat | 75.6 |
| Yi-6B-200K-GGUF | high | C-Eval | yi-6b-200k | TheBloke | 27 | 01-ai/Yi-6B-200K | 72 |
| Yi-6B-GGUF | high | C-Eval | yi-6b | TheBloke | 28 | 01-ai/Yi-6B | 72 |
| OrionStar-Yi-34B-Chat-Llama-GGUF | low | C-Eval | orionstar-yi-34b-chat-llama | TheBloke | 29 | OrionStarAI/OrionStar-Yi-34B-Chat-Llama | 69.9 |
| Qwen-14B-Chat | low | CEval (val) | qwen-14b-chat | Qwen | 30 | Qwen/Qwen-14B-Chat | 69.8 |
| Qwen-14B-Chat-GPTQ | low | CEval (val) | qwen-14b-chat | TheBloke | 31 | Qwen/Qwen-14B-Chat | 69.8 |
| Qwen-14B-Chat-AWQ | low | CEval (val) | qwen-14b-chat | TheBloke | 32 | Qwen/Qwen-14B-Chat | 69.8 |
| Qwen-14B-Chat-Int4 | high | CEval (val) | qwen-14b | Qwen | 33 | Qwen/Qwen-14B | 69 |
| Qwen-14B-Chat-Int8 | high | CEval (val) | qwen-14b | Qwen | 34 | Qwen/Qwen-14B | 68.6 |
| DeepSeek-V2-Chat | medium | **C-Eval** | deepseek-v2-chat | deepseek-ai | 35 | deepseek-ai/DeepSeek-V2-Chat | 67.5 |
| DeepSeek-V2 | medium | **C-Eval** | deepseek-v2 | deepseek-ai | 36 | deepseek-ai/DeepSeek-V2 | 67.5 |
| Nanbeige-16B-Base-GGUF | high | C-Eval | Unknown | TheBloke | 37 | Nanbeige/Nanbeige-16B-Base | 63.8 |
| Qwen2-72B-Instruct | medium | C-Eval | qwen2-72b | Qwen | 38 | Qwen/Qwen2-72B | 61.6 |
| Qwen-7B-Chat | low | CEval (val) | qwen-7b-chat | Qwen | 39 | Qwen/Qwen-7B-Chat | 59.7 |
| Qwen-7B-Chat-GPTQ | low | CEval (val) | qwen-7b-chat | TheBloke | 40 | Qwen/Qwen-7B-Chat | 59.7 |
| Qwen-7B-Chat-AWQ | low | CEval (val) | qwen-7b-chat | TheBloke | 41 | Qwen/Qwen-7B-Chat | 59.7 |
| Qwen-7B-Chat-Int8 | high | CEval (val) | qwen-7b | Qwen | 42 | Qwen/Qwen-7B | 59.4 |
| Qwen-7B-Chat-Int4 | high | CEval (val) | qwen-7b | Qwen | 43 | Qwen/Qwen-7B | 59.2 |
| Ring-lite | medium | C-Eval | ling-lite-base-1-5 | inclusionAI | 44 | inclusionAI/Ling-lite-base-1.5 | 59 |
| Yi-34B-GGUF | low | C-Eval | yi-34b | TheBloke | 45 | 01-ai/Yi-34B | 58.1 |
| Qwen-1_8B-Chat-Int8 | high | CEval (val) | qwen-1-8b | Qwen | 46 | Qwen/Qwen-1_8B | 55.8 |
| Llama3-70B-Chinese-Chat-GGUF-4bit | low | C-Eval Avg (Test Set) | meta-llama-3-70b | shenzhi-wang | 47 | meta-llama/Meta-Llama-3-70B-Instruct | 54.4 |
| Llama3-70B-Chinese-Chat | low | C-Eval Avg (Test Set) | meta-llama-3-70b | shenzhi-wang | 48 | meta-llama/Meta-Llama-3-70B-Instruct | 54.4 |
| internlm-chat-7b | medium | C-Eval(Val) | Unknown | internlm | 49 | internlm/internlm-chat-7b | 53.2 |
| internlm-7b | medium | C-Eval(Val) | Unknown | internlm | 50 | internlm/internlm-7b | 53.2 |
| Qwen-1_8B-Chat-Int4 | high | CEval (val) | qwen-1-8b | Qwen | 51 | Qwen/Qwen-1_8B | 52.8 |
| Skywork-13B-base-GGUF | medium | **C-EVAL** | Unknown | TheBloke | 52 | Skywork/Skywork-13B-base | 51.2 |
| Qwen2-7B | medium | C-Eval | qwen2-7b | Qwen | 53 | Qwen/Qwen2-7B | 47.4 |
| Qwen2-7B-Instruct | medium | C-Eval | qwen2-7b | Qwen | 54 | Qwen/Qwen2-7B | 45.9 |
| DeepSeek-V2-Lite-Chat | medium | **C-Eval** | deepseek-v2-lite-chat | deepseek-ai | 55 | deepseek-ai/DeepSeek-V2-Lite-Chat | 45 |
| DeepSeek-V2-Lite | medium | **C-Eval** | deepseek-v2-lite | deepseek-ai | 56 | deepseek-ai/DeepSeek-V2-Lite | 45 |
| Qwen2-1.5B-Instruct | medium | C-Eval | qwen2-1-5b-instruct | Qwen | 57 | Qwen/Qwen2-1.5B-Instruct | 37.2 |
| Qwen2-0.5B-Instruct | medium | C-Eval | qwen2-0-5b | Qwen | 58 | Qwen/Qwen2-0.5B | 37.2 |
| Skywork-13B-base-GGUF | low | C-Eval | Unknown | TheBloke | 59 | Skywork/Skywork-13B-base | 35.5 |
| Qwen-72B | low | C-Eval | qwen-72b | Qwen | 60 | Qwen/Qwen-72B | 32.5 |
| Qwen-7B | low | C-Eval | qwen-7b | Qwen | 61 | Qwen/Qwen-7B | 32.5 |
| Qwen-14B | low | C-Eval | qwen-14b | Qwen | 62 | Qwen/Qwen-14B | 32.5 |
| Qwen2-0.5B | medium | C-Eval | qwen2-0-5b | Qwen | 63 | Qwen/Qwen2-0.5B | 23.4 |
| Qwen2-1.5B | medium | C-Eval | qwen2-1-5b | Qwen | 64 | Qwen/Qwen2-1.5B | 23.4 |
Provenance & metadata (1)
source
- kind
- leaderboard-scrape
- paths
- benchmarks/ceval.html, docs/notes/hf-benchmark-contexts.json