OpenCompass LLM
Indexed results from OpenCompass LLM. Each ranking retains its own metric and evaluation conditions.
Average
openCompass · OpenCompass LLM · Average · percent · Higher is better
- Z.ai: GLM 5z-ai79%
- Qwen: Qwen3.5 397B A17Bqwen78.4%
- deepseek-ai/DeepSeek-V3.2-Specialedeepseek78.2%
- Z.ai: GLM 4.7z-ai77.6%
- StepFun: Step 3.5 Flashstepfun76.9%
- DeepSeek: DeepSeek V3.2deepseek75.3%
- MoonshotAI: Kimi K2 Thinkingmoonshotai75.3%
- MiniMax: MiniMax M2.5minimax73.3%
- Z.ai: GLM 4.6z-ai73.3%
- OpenAI: gpt-oss-120bopenai73.2%
- Qwen: Qwen3 235B A22B Thinking 2507qwen71.8%
- MiniMax: MiniMax M2minimax69.5%
Average
openCompass · OpenCompass LLM · Average · percent · Higher is better
Average
openCompass · OpenCompass LLM · Average · percent · Higher is better
IFEval
openCompass · OpenCompass LLM · IFEval · percent · Higher is better
- StepFun: Step 3.5 Flashstepfun93.2%
- Z.ai: GLM 5z-ai93.2%
- MoonshotAI: Kimi K2 Thinkingmoonshotai92.4%
- deepseek-ai/DeepSeek-V3.2-Specialedeepseek91.7%
- Qwen: Qwen3.5 397B A17Bqwen91.5%
- MiniMax: MiniMax M2.5minimax91.1%
- MiniMax: MiniMax M2minimax90.2%
- OpenAI: gpt-oss-120bopenai90.2%
- Z.ai: GLM 4.7z-ai90.2%
- DeepSeek: DeepSeek V3.2deepseek89.7%
- Qwen: Qwen3 30B A3B Thinking 2507qwen89.7%
- Qwen: Qwen3 Next 80B A3B Thinkingqwen89.5%
LiveCodeBenchV6
openCompass · OpenCompass LLM · LiveCodeBenchV6 · percent · Higher is better
- Z.ai: GLM 5z-ai86.2%
- StepFun: Step 3.5 Flashstepfun83.9%
- Z.ai: GLM 4.7z-ai83.8%
- Qwen: Qwen3.5 397B A17Bqwen83%
- deepseek-ai/DeepSeek-V3.2-Specialedeepseek80.9%
- OpenAI: gpt-oss-120bopenai78.4%
- Z.ai: GLM 4.6z-ai78.2%
- MoonshotAI: Kimi K2 Thinkingmoonshotai77.1%
- DeepSeek: DeepSeek V3.2deepseek75.4%
- MiniMax: MiniMax M2minimax74%
- MiniMax: MiniMax M2.5minimax73.6%
- Qwen: Qwen3 235B A22B Thinking 2507qwen70.6%
Average
openCompass · OpenCompass LLM · Average · percent · Higher is better
Academic_Reasoning
openCompass · OpenCompass LLM · Academic_Reasoning · percent · Higher is better
Math
openCompass · OpenCompass LLM · Math · percent · Higher is better
GPQA-Diamond
openCompass · OpenCompass LLM · GPQA-Diamond · percent · Higher is better
- Qwen: Qwen3.5 397B A17Bqwen88.4%
- Z.ai: GLM 4.7z-ai86.9%
- deepseek-ai/DeepSeek-V3.2-Specialedeepseek86.7%
- Z.ai: GLM 5z-ai85.3%
- DeepSeek: DeepSeek V3.2deepseek84.6%
- MiniMax: MiniMax M2.5minimax84.6%
- StepFun: Step 3.5 Flashstepfun83.7%
- MoonshotAI: Kimi K2 Thinkingmoonshotai82.7%
- DeepSeek: R1 0528deepseek80.6%
- Z.ai: GLM 4.6z-ai80.4%
- Qwen: Qwen3 235B A22B Thinking 2507qwen79.8%
- Z.ai: GLM 4.5z-ai79.5%
Knowledge_Science
openCompass · OpenCompass LLM · Knowledge_Science · percent · Higher is better
MMLU-Pro
openCompass · OpenCompass LLM · MMLU-Pro · percent · Higher is better
- Qwen: Qwen3.5 397B A17Bqwen87.6%
- DeepSeek: DeepSeek V3.2deepseek85.8%
- deepseek-ai/DeepSeek-V3.2-Specialedeepseek85.5%
- Z.ai: GLM 5z-ai85.2%
- MoonshotAI: Kimi K2 Thinkingmoonshotai84.3%
- Z.ai: GLM 4.7z-ai84%
- DeepSeek: R1 0528deepseek83.5%
- Qwen: Qwen3 235B A22B Thinking 2507qwen83.5%
- StepFun: Step 3.5 Flashstepfun83.5%
- Z.ai: GLM 4.6z-ai83%
- Z.ai: GLM 4.5z-ai82.7%
- Qwen: Qwen3 Next 80B A3B Thinkingqwen82%
HLE
openCompass · OpenCompass LLM · HLE · percent · Higher is better
- deepseek-ai/DeepSeek-V3.2-Specialedeepseek28.6%
- Z.ai: GLM 5z-ai28.1%
- Qwen: Qwen3.5 397B A17Bqwen27.5%
- Z.ai: GLM 4.7z-ai25.4%
- DeepSeek: DeepSeek V3.2deepseek23.2%
- MiniMax: MiniMax M2.5minimax22.2%
- StepFun: Step 3.5 Flashstepfun21.6%
- MoonshotAI: Kimi K2 Thinkingmoonshotai21.3%
- Z.ai: GLM 4.6z-ai19.3%
- Qwen: Qwen3 235B A22B Thinking 2507qwen18.5%
- OpenAI: gpt-oss-120bopenai18.3%
- Z.ai: GLM 4.5z-ai16.9%
Models
LoadingBenchmarks
Explore every benchmark
LoadingThese charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.