Model cards
What 60 AI models are reported to be good and bad at, with the source and date for every claim.
| Model | Vendor | Tier | Price in / out | Coding | Agentic coding | Sources | Latest source |
|---|---|---|---|---|---|---|---|
| qwen3-coder-480b | alibaba | - | - | - | 0.33 | 1 / 0 / 0 | 2026-09-14 |
| qwen3.8-27b | alibaba | - | - | 0.04 | - | 2 / 0 / 0 | 2026-10-19 |
| qwen3.8-flash-next | alibaba | - | - | 0.26 | - | 1 / 0 / 0 | - |
| qwen3.8-max | alibaba | - | $2 / $6 | 0.73 | 0.55 | 9 / 11 / 9 | 2026-10-19 |
| claude-fable-5 | anthropic | frontier | $10 / $50 | 0.17 | 0.19 | 2 / 0 / 0 | 2026-10-20 |
| claude-fable-5.1 | anthropic | frontier | $10 / $50 | 0.83 | 0.55 | 12 / 12 / 16 | 2026-10-19 |
| claude-haiku-5.5 | anthropic | fast | $0.1 / $0.5 | - | 0.46 | 7 / 13 / 10 | 2026-10-18 |
| claude-mythos-5 | anthropic | frontier | $10 / $50 | - | 0.71 | 1 / 0 / 0 | 2026-10-19 |
| claude-mythos-5.1 | anthropic | frontier | $10 / $50 | - | 0.88 | 1 / 0 / 2 | 2026-10-05 |
| claude-opus-4.5 | anthropic | deep | $5 / $25 | - | 0.67 | 1 / 0 / 0 | 2026-09-14 |
| claude-opus-4.6 | anthropic | deep | $5 / $25 | - | 0.80 | 1 / 0 / 0 | 2026-09-14 |
| claude-opus-5 | anthropic | deep | $5 / $25 | 0.74 | 0.66 | 6 / 1 / 3 | 2026-10-19 |
| claude-opus-5.5 | anthropic | deep | $4 / $20 | 1.00 | 0.85 | 10 / 17 / 22 | 2026-10-19 |
| claude-sonnet-4.5 | anthropic | deep | $3 / $15 | - | 0.60 | 1 / 0 / 0 | 2026-09-14 |
| claude-sonnet-5 | anthropic | standard | $2 / $10 | - | 0.28 | 6 / 4 / 3 | 2026-09-28 |
| claude-sonnet-5.5 | anthropic | standard | $2 / $10 | 0.91 | 0.62 | 12 / 16 / 16 | 2026-10-20 |
| deepseek-v3.2 | deepseek | - | - | - | 0.07 | 1 / 0 / 0 | 2026-09-14 |
| deepseek-v4-flash-vision-exp | deepseek | - | - | - | 0.13 | 1 / 0 / 0 | 2026-10-20 |
| deepseek-v4-pro | deepseek | - | - | - | 0.54 | 5 / 5 / 4 | 2026-10-19 |
| deepseek-v4.1-flash | deepseek | - | - | 0.57 | 0.83 | 8 / 14 / 8 | 2026-10-20 |
| gemini-3-flash | - | $0.5 / $3 | - | 0.14 | 7 / 0 / 1 | 2026-09-14 | |
| gemini-3-pro | - | - | - | 0.53 | 1 / 0 / 0 | 2026-09-14 | |
| gemini-3.1-pro | - | $2 / $12 | - | 0.43 | 6 / 9 / 8 | 2026-10-19 | |
| gemini-3.5-flash-lite | - | $0.3 / $2.5 | - | 0.06 | 6 / 9 / 4 | 2026-10-19 | |
| gemini-3.7-flash | - | - | 0.00 | 0.42 | 2 / 0 / 0 | 2026-10-19 | |
| gemini-3.8-flash | - | $0.75 / $3.75 | - | 0.64 | 6 / 12 / 15 | 2026-10-19 | |
| gemini-4-argon | - | $2 / $10 | 0.68 | 0.88 | 6 / 14 / 13 | 2026-10-19 | |
| muse-spark | meta | - | - | - | 0.87 | 1 / 0 / 0 | 2026-09-14 |
| muse-spark-1.1 | meta | - | - | - | 1.00 | 1 / 0 / 0 | 2026-09-14 |
| muse-spark-1.3 | meta | - | - | 0.52 | 0.85 | 3 / 0 / 0 | 2026-10-20 |
| muse-wonder | meta | - | - | 0.43 | - | 2 / 0 / 0 | 2026-10-20 |
| muse-wonder-full | meta | - | - | 0.89 | - | 3 / 5 / 1 | 2026-10-20 |
| devstral-2 | mistral | - | - | - | 0.00 | 1 / 1 / 0 | 2025-12-09 |
| mistral-large-4 | mistral | - | $0.68 / $2.09 | - | 0.09 | 3 / 10 / 8 | 2026-10-19 |
| kimi-k2 | moonshot | - | - | - | 0.13 | 1 / 0 / 0 | 2026-09-14 |
| kimi-k3 | moonshot | - | $3 / $15 | 0.45 | 0.76 | 7 / 10 / 10 | 2026-10-19 |
| kimi-k3.1-pro | moonshot | - | - | - | - | 1 / 5 / 0 | 2026-10-26 |
| gpt-5 | openai | standard | $1.25 / $10 | - | 0.47 | 1 / 0 / 0 | 2026-09-14 |
| gpt-5.2 | openai | standard | $1.75 / $14 | - | 0.20 | 1 / 0 / 0 | 2026-09-14 |
| gpt-5.2-codex | openai | standard | $1.75 / $14 | - | 0.40 | 1 / 0 / 0 | 2026-09-14 |
| gpt-5.3-codex | openai | standard | $1.75 / $14 | - | 0.33 | 2 / 0 / 1 | 2026-07-12 |
| gpt-5.4 | openai | standard | $2.5 / $15 | - | 0.93 | 1 / 0 / 0 | 2026-09-14 |
| gpt-5.5 | openai | deep | $5 / $30 | - | 0.42 | 2 / 2 / 4 | 2026-09-21 |
| gpt-5.6-cyber | openai | frontier | $12.5 / $75 | - | - | 0 / 0 / 4 | 2026-08-13 |
| gpt-5.6-luna | openai | fast | $0.2 / $1.2 | - | 0.45 | 2 / 0 / 4 | 2026-10-20 |
| gpt-5.6-sol | openai | deep | $4 / $20 | - | 0.70 | 6 / 9 / 12 | 2026-10-19 |
| gpt-5.6-terra | openai | standard | $2 / $12 | - | 0.67 | 2 / 0 / 6 | 2026-10-19 |
| gpt-6-astra | openai | frontier | $10 / $50 | 0.96 | 0.55 | 9 / 14 / 14 | 2026-10-19 |
| gpt-6-luna | openai | fast | $0.1 / $0.5 | - | 0.08 | 6 / 11 / 7 | 2026-10-08 |
| gpt-6-sol | openai | standard | $2 / $10 | 0.70 | 0.20 | 9 / 4 / 7 | 2026-10-20 |
| gpt-6.1-sol | openai | standard | $2 / $10 | 0.87 | 0.74 | 8 / 16 / 15 | 2026-10-19 |
| hy4-preview | tencent | - | - | 0.30 | 0.38 | 2 / 0 / 0 | 2026-10-20 |
| grok-4.5 | xai | - | - | - | 0.08 | 1 / 0 / 0 | 2026-10-20 |
| grok-4.7 | xai | - | $2 / $6 | 0.34 | 0.26 | 7 / 5 / 1 | 2026-10-20 |
| mimo-v2.6-flash | xiaomi | - | - | 0.35 | 0.60 | 2 / 0 / 0 | 2026-10-20 |
| mimo-v2.6-pro | xiaomi | - | - | 0.22 | 0.96 | 3 / 0 / 0 | 2026-10-20 |
| glm-4.6 | zhipu | - | - | - | 0.00 | 1 / 0 / 0 | 2026-09-14 |
| glm-5.2 | zhipu | - | $1.4 / $4.4 | 0.04 | 0.42 | 6 / 8 / 12 | 2026-07-08 |
| glm-5.3 | zhipu | - | - | 0.14 | 0.53 | 6 / 4 / 2 | 2026-10-20 |
| glm-5.3-flash | zhipu | - | - | - | 0.19 | 1 / 0 / 0 | 2026-10-20 |
Percentiles compare a model with the others listed here, on benchmarks at least four of them share. A dash means no data, not a low score. Tier places a model within its vendor's lineup by price.