All models
Every model family with community-reported local performance, grouped by publisher.
Alibaba
Google DeepMind
DeepSeek
Zhipu AI
Meta
OpenAI
PrismML
Moonshot AI
Xiaomi
Liquid AI
NVIDIA
Ant Group
Microsoft
Kwaipilot
MiniMax
InternScience
StepFun
Tencent
Other
- DeepSeek V4.1 Flash52
- Llama 231
- Ornith1.527
- Qwen3-Coder25
- Qwen 2.5 Coder9
- Nemotron 3.57
- Qwen3-Coder-Next7
- DeepSeek R1 Distill6
- LFM36
- Nemotron 36
- GLM-4.7-Flash5
- Minimax M2.15
- GLM-4.5-Air4
- Mellum2.14
- Qwen3-VL4
- DeepSeek V4.13
- Falcon-H13
- MiniCPM-V 4.73
- Nemotron 3 Nano3
- Step-3.5-Flash3
- DeepSeek V22
- Edge02
- Gemma 22
- Granite-4.0-H2
- Llama 32
- Llama22
- MiniMax M2.72
- MiniMax M3.12
- Mistral Small 3.12
- Nemotron 3.5 Super2
- Nex-N2.5-mini2
- Ornith 1.02
- Poolside Laguna S 2.12
- Qwen2.5-Omni2
- Raon-12
- Shisa V2 Llama 3.32
- SmolLM32
- BitNet b1.581
- DeepSeek V3.11
- DeepSeek-Coder-V2-Lite1
- Falcon31
- GLM-41
- GLM-4.6V1
- Granite 3.21
- Granite Switch 4.11
- HobbyLM1
- Laguna S 2.11
- LFM21
- LLaDA-2.1-mini1
- Llama 4 Scout1
- MiniCPM51
- MiniMax M2.51
- Ministral 31
- Mistral-8x22b1
- Nanbeige4.31
- Nemotron 3.5 Bolt1
- Nemotron-Labs Dory 1.01
- Phi-31
- Qwen21
- Qwopus3.61
- RWKV71
- SenseNova-U1.51
- SmolLM21
- Step 3.7 Flash1