vals-style side-by-side comparison · max 4 modelsCOMPARECompare modelsSee how models stack up on every benchmark — accuracy, latency, cost, plus a radar chart of category strengths. 📋 COPY URL RESETMODEL 1 — pick a model —claude-cli/claude-fableclaude-cli/claude-fable-4claude-cli/claude-fable-5claude-cli/claude-fable-5-tclaude-cli/claude-haikuclaude-cli/claude-opusclaude-cli/claude-opus-4-8claude-cli/claude-sonnetclaude-cli/claude-sonnet-4-6claude-fableclaude-haikuclaude-opusclaude-opus-5claude-opus-5-highclaude-sonnetcodex-gpt-5.3-sparkcodex-gpt-5.4codex-gpt-5.4-minicodex-gpt-5.5codex-gpt-5.6-lunacodex-gpt-5.6-solcodex-gpt-5.6-terragemini-3.1-pro-highgemini-3.6-flash-highgrok-3-mini-directgrok-4-1-fast-directgrok-4-fast-directgrok-4.20-non-reasoning-directgrok-4.20-reasoning-directgrok-4.3-directgrok-4.3-latest-directgrok-4.5-directgrok-4.6-directgrok-build-0.1kimi-k3minimax-m2.7minimax-m2.7-highspeedminimax-m3minimax-mediummistral-medium-2508mistral/devstralmistral/magistral-mediummistral/magistral-smallmistral/mistral-largemistral/mistral-mediummuse-glimmer-30b-q4_k_xl-localoai/grok-4-1-fast-reasoningollama-cl/deepseek-v4-flashollama-cl/deepseek-v4-proollama-cl/gemma4ollama-cl/gemma4:26bollama-cl/gpt-oss-120bollama-cl/gpt-oss:120bollama-cl/gpt-oss:20bollama-cl/kimi-k2.5ollama-cl/kimi-k2.6ollama-cl/kimi-k2.7-codeollama-cl/minimax-m2.5ollama-cl/mistral-large-3-675bollama-cl/mistral-large-3:675bollama-cl/nemotron-3-nano-30bollama-cl/nemotron-3-nano:30bollama-cl/nemotron-3-superollama-cl/nemotron-3-ultraollama-cl/qwen3-coder-nextollama-cl/qwen3-coder:480bollama-cl/qwen3.5ollama-cl/qwen3.5-codeollama-cl/qwen3.5:27bollama-cl/qwen3.5:397bxai/grok-3-minixai/grok-4-1-fast-reasoningxai/grok-4.3-latestxai/grok-4.5zai-coding/glm-4.6zai-coding/glm-4.7zai-coding/glm-5zai-coding/glm-5-turbozai-coding/glm-5.1zai-coding/glm-5.2zai-coding/glm-5.3-flashMODEL 2 — pick a model —claude-cli/claude-fableclaude-cli/claude-fable-4claude-cli/claude-fable-5claude-cli/claude-fable-5-tclaude-cli/claude-haikuclaude-cli/claude-opusclaude-cli/claude-opus-4-8claude-cli/claude-sonnetclaude-cli/claude-sonnet-4-6claude-fableclaude-haikuclaude-opusclaude-opus-5claude-opus-5-highclaude-sonnetcodex-gpt-5.3-sparkcodex-gpt-5.4codex-gpt-5.4-minicodex-gpt-5.5codex-gpt-5.6-lunacodex-gpt-5.6-solcodex-gpt-5.6-terragemini-3.1-pro-highgemini-3.6-flash-highgrok-3-mini-directgrok-4-1-fast-directgrok-4-fast-directgrok-4.20-non-reasoning-directgrok-4.20-reasoning-directgrok-4.3-directgrok-4.3-latest-directgrok-4.5-directgrok-4.6-directgrok-build-0.1kimi-k3minimax-m2.7minimax-m2.7-highspeedminimax-m3minimax-mediummistral-medium-2508mistral/devstralmistral/magistral-mediummistral/magistral-smallmistral/mistral-largemistral/mistral-mediummuse-glimmer-30b-q4_k_xl-localoai/grok-4-1-fast-reasoningollama-cl/deepseek-v4-flashollama-cl/deepseek-v4-proollama-cl/gemma4ollama-cl/gemma4:26bollama-cl/gpt-oss-120bollama-cl/gpt-oss:120bollama-cl/gpt-oss:20bollama-cl/kimi-k2.5ollama-cl/kimi-k2.6ollama-cl/kimi-k2.7-codeollama-cl/minimax-m2.5ollama-cl/mistral-large-3-675bollama-cl/mistral-large-3:675bollama-cl/nemotron-3-nano-30bollama-cl/nemotron-3-nano:30bollama-cl/nemotron-3-superollama-cl/nemotron-3-ultraollama-cl/qwen3-coder-nextollama-cl/qwen3-coder:480bollama-cl/qwen3.5ollama-cl/qwen3.5-codeollama-cl/qwen3.5:27bollama-cl/qwen3.5:397bxai/grok-3-minixai/grok-4-1-fast-reasoningxai/grok-4.3-latestxai/grok-4.5zai-coding/glm-4.6zai-coding/glm-4.7zai-coding/glm-5zai-coding/glm-5-turbozai-coding/glm-5.1zai-coding/glm-5.2zai-coding/glm-5.3-flash+ ADD 3RDMODEL 3 — pick a model —claude-cli/claude-fableclaude-cli/claude-fable-4claude-cli/claude-fable-5claude-cli/claude-fable-5-tclaude-cli/claude-haikuclaude-cli/claude-opusclaude-cli/claude-opus-4-8claude-cli/claude-sonnetclaude-cli/claude-sonnet-4-6claude-fableclaude-haikuclaude-opusclaude-opus-5claude-opus-5-highclaude-sonnetcodex-gpt-5.3-sparkcodex-gpt-5.4codex-gpt-5.4-minicodex-gpt-5.5codex-gpt-5.6-lunacodex-gpt-5.6-solcodex-gpt-5.6-terragemini-3.1-pro-highgemini-3.6-flash-highgrok-3-mini-directgrok-4-1-fast-directgrok-4-fast-directgrok-4.20-non-reasoning-directgrok-4.20-reasoning-directgrok-4.3-directgrok-4.3-latest-directgrok-4.5-directgrok-4.6-directgrok-build-0.1kimi-k3minimax-m2.7minimax-m2.7-highspeedminimax-m3minimax-mediummistral-medium-2508mistral/devstralmistral/magistral-mediummistral/magistral-smallmistral/mistral-largemistral/mistral-mediummuse-glimmer-30b-q4_k_xl-localoai/grok-4-1-fast-reasoningollama-cl/deepseek-v4-flashollama-cl/deepseek-v4-proollama-cl/gemma4ollama-cl/gemma4:26bollama-cl/gpt-oss-120bollama-cl/gpt-oss:120bollama-cl/gpt-oss:20bollama-cl/kimi-k2.5ollama-cl/kimi-k2.6ollama-cl/kimi-k2.7-codeollama-cl/minimax-m2.5ollama-cl/mistral-large-3-675bollama-cl/mistral-large-3:675bollama-cl/nemotron-3-nano-30bollama-cl/nemotron-3-nano:30bollama-cl/nemotron-3-superollama-cl/nemotron-3-ultraollama-cl/qwen3-coder-nextollama-cl/qwen3-coder:480bollama-cl/qwen3.5ollama-cl/qwen3.5-codeollama-cl/qwen3.5:27bollama-cl/qwen3.5:397bxai/grok-3-minixai/grok-4-1-fast-reasoningxai/grok-4.3-latestxai/grok-4.5zai-coding/glm-4.6zai-coding/glm-4.7zai-coding/glm-5zai-coding/glm-5-turbozai-coding/glm-5.1zai-coding/glm-5.2zai-coding/glm-5.3-flash+ ADD 4THMODEL 4 — pick a model —claude-cli/claude-fableclaude-cli/claude-fable-4claude-cli/claude-fable-5claude-cli/claude-fable-5-tclaude-cli/claude-haikuclaude-cli/claude-opusclaude-cli/claude-opus-4-8claude-cli/claude-sonnetclaude-cli/claude-sonnet-4-6claude-fableclaude-haikuclaude-opusclaude-opus-5claude-opus-5-highclaude-sonnetcodex-gpt-5.3-sparkcodex-gpt-5.4codex-gpt-5.4-minicodex-gpt-5.5codex-gpt-5.6-lunacodex-gpt-5.6-solcodex-gpt-5.6-terragemini-3.1-pro-highgemini-3.6-flash-highgrok-3-mini-directgrok-4-1-fast-directgrok-4-fast-directgrok-4.20-non-reasoning-directgrok-4.20-reasoning-directgrok-4.3-directgrok-4.3-latest-directgrok-4.5-directgrok-4.6-directgrok-build-0.1kimi-k3minimax-m2.7minimax-m2.7-highspeedminimax-m3minimax-mediummistral-medium-2508mistral/devstralmistral/magistral-mediummistral/magistral-smallmistral/mistral-largemistral/mistral-mediummuse-glimmer-30b-q4_k_xl-localoai/grok-4-1-fast-reasoningollama-cl/deepseek-v4-flashollama-cl/deepseek-v4-proollama-cl/gemma4ollama-cl/gemma4:26bollama-cl/gpt-oss-120bollama-cl/gpt-oss:120bollama-cl/gpt-oss:20bollama-cl/kimi-k2.5ollama-cl/kimi-k2.6ollama-cl/kimi-k2.7-codeollama-cl/minimax-m2.5ollama-cl/mistral-large-3-675bollama-cl/mistral-large-3:675bollama-cl/nemotron-3-nano-30bollama-cl/nemotron-3-nano:30bollama-cl/nemotron-3-superollama-cl/nemotron-3-ultraollama-cl/qwen3-coder-nextollama-cl/qwen3-coder:480bollama-cl/qwen3.5ollama-cl/qwen3.5-codeollama-cl/qwen3.5:27bollama-cl/qwen3.5:397bxai/grok-3-minixai/grok-4-1-fast-reasoningxai/grok-4.3-latestxai/grok-4.5zai-coding/glm-4.6zai-coding/glm-4.7zai-coding/glm-5zai-coding/glm-5-turbozai-coding/glm-5.1zai-coding/glm-5.2zai-coding/glm-5.3-flashAVG. ACCURACY AVG. LATENCY AVG. COSTMETRICPick two or more models above. Default shows the top 2 from your fleet.BenchmarksAccuracy · RankingsPerformance breakdownCost analysisBars show 1P-estimated cost per million tokens for input and output.
