All models
77 models, from 16 makers. Search by name, filter by maker, or show only open-weight models.
Showing all 77 models. Filtering needs JavaScript, which runs only in your browser.
- Top tier
- Strong
- Middle
- Behind the leaders
- Not enough data
Ratings compare only the 77 models listed here, in quarters, best first. “Not enough data” means no result yet. It is never a low score.
- WritingBehind the leadersCodingMiddleMathsNot enough dataKnowledgeNot enough data
- Claude Fable 5, see the evidenceAnthropicWritingTop tierCodingTop tierMathsTop tierKnowledgeTop tier
- Claude Fable 5.1, see the evidenceAnthropicWritingTop tierCodingTop tierMathsTop tierKnowledgeTop tier
- Claude Haiku 5.5, see the evidenceAnthropicWritingBehind the leadersCodingBehind the leadersMathsStrongKnowledgeBehind the leaders
- WritingStrongCodingStrongMathsNot enough dataKnowledgeNot enough data
- Claude Opus 4.6, see the evidenceAnthropicWritingTop tierCodingTop tierMathsMiddleKnowledgeMiddle
- Claude Opus 4.7, see the evidenceAnthropicWritingTop tierCodingTop tierMathsStrongKnowledgeStrong
- Claude Opus 4.8, see the evidenceAnthropicWritingStrongCodingStrongMathsStrongKnowledgeStrong
- Claude Opus 5, see the evidenceAnthropicWritingTop tierCodingTop tierMathsTop tierKnowledgeStrong
- Claude Opus 5.5, see the evidenceAnthropicWritingTop tierCodingTop tierMathsTop tierKnowledgeTop tier
- Claude Sonnet 4.6, see the evidenceAnthropicWritingMiddleCodingTop tierMathsNot enough dataKnowledgeBehind the leaders
- Claude Sonnet 5, see the evidenceAnthropicWritingBehind the leadersCodingMiddleMathsMiddleKnowledgeBehind the leaders
- Claude Sonnet 5.5, see the evidenceAnthropicWritingTop tierCodingTop tierMathsTop tierKnowledgeMiddle
- DeepSeek-V4-Pro, see the evidenceDeepSeekOpen weightsWritingStrongCodingMiddleMathsBehind the leadersKnowledgeMiddle
- DeepSeek-V4-Pro (High, 20260813), see the evidenceDeepSeekOpen weightsWritingStrongCodingMiddleMathsMiddleKnowledgeStrong
- DeepSeek-V4-Pro Preview, see the evidenceDeepSeekOpen weightsWritingStrongCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- DeepSeek-V4.1-Flash-20260910, see the evidenceDeepSeekOpen weightsWritingMiddleCodingTop tierMathsNot enough dataKnowledgeNot enough data
- Dola Seed 2.0 Pro, see the evidenceByteDanceWritingBehind the leadersCodingMiddleMathsNot enough dataKnowledgeNot enough data
- WritingMiddleCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingStrongCodingStrongMathsNot enough dataKnowledgeNot enough data
- Gemini 2.5 Pro, see the evidenceGoogle DeepMindWritingStrongCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- Gemini 3 Flash, see the evidenceGoogle DeepMindWritingStrongCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- Gemini 3 Pro, see the evidenceGoogle DeepMindWritingTop tierCodingMiddleMathsNot enough dataKnowledgeNot enough data
- Gemini 3.1 Pro Preview, see the evidenceGoogle DeepMindWritingTop tierCodingMiddleMathsNot enough dataKnowledgeNot enough data
- Gemini 3.5 Flash, see the evidenceGoogle DeepMindWritingTop tierCodingStrongMathsMiddleKnowledgeTop tier
- Gemini 3.5 Flash-Lite, see the evidenceGoogle DeepMindWritingMiddleCodingBehind the leadersMathsBehind the leadersKnowledgeNot enough data
- Gemini 3.6 Flash, see the evidenceGoogle DeepMindWritingTop tierCodingStrongMathsMiddleKnowledgeTop tier
- Gemini 3.7 Flash, see the evidenceGoogle DeepMindWritingTop tierCodingTop tierMathsStrongKnowledgeTop tier
- Gemini 3.8 Flash, see the evidenceGoogle DeepMindWritingTop tierCodingTop tierMathsMiddleKnowledgeTop tier
- Gemini 4 Argon, see the evidenceGoogle DeepMindWritingTop tierCodingTop tierMathsTop tierKnowledgeTop tier
- Gemma 4 31B, see the evidenceGoogle DeepMindOpen weightsWritingBehind the leadersCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- GLM-4.7, see the evidenceZ.ai (Zhipu AI)Open weightsWritingBehind the leadersCodingBehind the leadersMathsNot enough dataKnowledgeBehind the leaders
- GLM-5, see the evidenceZ.ai (Zhipu AI)Open weightsWritingMiddleCodingMiddleMathsNot enough dataKnowledgeNot enough data
- GLM-5.1, see the evidenceZ.ai (Zhipu AI)Open weightsWritingStrongCodingStrongMathsBehind the leadersKnowledgeBehind the leaders
- GLM-5.2, see the evidenceZ.ai (Zhipu AI)Open weightsWritingTop tierCodingStrongMathsMiddleKnowledgeBehind the leaders
- GLM-5.3, see the evidenceZ.ai (Zhipu AI)Open weightsWritingStrongCodingStrongMathsStrongKnowledgeMiddle
- GLM-5.3-Flash, see the evidenceZ.ai (Zhipu AI)Open weightsWritingStrongCodingTop tierMathsBehind the leadersKnowledgeNot enough data
- WritingMiddleCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingStrongCodingStrongMathsNot enough dataKnowledgeNot enough data
- WritingStrongCodingStrongMathsStrongKnowledgeStrong
- WritingMiddleCodingBehind the leadersMathsBehind the leadersKnowledgeNot enough data
- WritingBehind the leadersCodingMiddleMathsStrongKnowledgeMiddle
- WritingStrongCodingStrongMathsTop tierKnowledgeTop tier
- WritingBehind the leadersCodingStrongMathsTop tierKnowledgeMiddle
- WritingMiddleCodingStrongMathsTop tierKnowledgeTop tier
- WritingBehind the leadersCodingBehind the leadersMathsStrongKnowledgeMiddle
- WritingBehind the leadersCodingBehind the leadersMathsTop tierKnowledgeStrong
- WritingMiddleCodingStrongMathsTop tierKnowledgeTop tier
- WritingMiddleCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingStrongCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingMiddleCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingBehind the leadersCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingStrongCodingMiddleMathsMiddleKnowledgeMiddle
- WritingMiddleCodingMiddleMathsMiddleKnowledgeMiddle
- WritingBehind the leadersCodingBehind the leadersMathsBehind the leadersKnowledgeStrong
- Inkling-Small, see the evidenceThinking MachinesOpen weightsWritingBehind the leadersCodingBehind the leadersMathsBehind the leadersKnowledgeBehind the leaders
- Kimi K2.5, see the evidenceMoonshotOpen weightsWritingMiddleCodingMiddleMathsNot enough dataKnowledgeBehind the leaders
- Kimi K2.6, see the evidenceMoonshotOpen weightsWritingMiddleCodingStrongMathsMiddleKnowledgeBehind the leaders
- Kimi K3, see the evidenceMoonshotOpen weightsWritingStrongCodingTop tierMathsStrongKnowledgeStrong
- MiMo-V2.5-Pro, see the evidenceXiaomiOpen weightsWritingMiddleCodingTop tierMathsNot enough dataKnowledgeNot enough data
- MiMo-V2.6-Flash, see the evidenceXiaomiOpen weightsWritingBehind the leadersCodingStrongMathsNot enough dataKnowledgeNot enough data
- MiMo-V2.6-Pro, see the evidenceXiaomiOpen weightsWritingTop tierCodingTop tierMathsNot enough dataKnowledgeNot enough data
- Mistral Large 4, see the evidenceMistral AIWritingBehind the leadersCodingMiddleMathsMiddleKnowledgeBehind the leaders
- WritingStrongCodingMiddleMathsNot enough dataKnowledgeNot enough data
- WritingMiddleCodingStrongMathsNot enough dataKnowledgeStrong
- WritingMiddleCodingTop tierMathsNot enough dataKnowledgeStrong
- WritingTop tierCodingTop tierMathsStrongKnowledgeNot enough data
- WritingTop tierCodingTop tierMathsNot enough dataKnowledgeNot enough data
- NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4), see the evidenceNVIDIAOpen weightsWritingBehind the leadersCodingMiddleMathsNot enough dataKnowledgeNot enough data
- Qwen3.5-397B-A17B, see the evidenceAlibabaOpen weightsWritingBehind the leadersCodingMiddleMathsBehind the leadersKnowledgeNot enough data
- WritingBehind the leadersCodingBehind the leadersMathsNot enough dataKnowledgeNot enough data
- WritingTop tierCodingStrongMathsNot enough dataKnowledgeNot enough data
- WritingMiddleCodingMiddleMathsNot enough dataKnowledgeStrong
- WritingBehind the leadersCodingMiddleMathsBehind the leadersKnowledgeMiddle
- WritingStrongCodingStrongMathsNot enough dataKnowledgeNot enough data
- WritingMiddleCodingMiddleMathsBehind the leadersKnowledgeNot enough data
- WritingTop tierCodingTop tierMathsStrongKnowledgeMiddle
No model matches. Clear the search or pick another maker.
Listed alphabetically, not by rank. “Open weights” comes from the licence Arena lists (anything other than Proprietary) or from Epoch AI's access label. It is not a price: it does not tell you whether an app is free. See models, apps and harnesses.
Models held back for review: 1. We do not publish results for a model until its identifiers across sources have been checked. See open issues.