Compare models head-to-head
Pick one or two models (A and B) on the left. The panel compares their benchmark scores and cheapest 10:1 price prominently — the better value in each row is highlighted — and lists the cheapest providers for each, side by side. Filters at the top (score, providers, models) apply here too.
Click to pick up to two models (A then B). Click again to deselect.
| A/B | Model | Score |
|---|---|---|
| Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | 96.1 | |
| Kimi K3 (max)open | 95.2 | |
| Claude Opus 5 (Adaptive Reasoning, High Effort) | 94.1 | |
| Grok 4.6 (high) | 93.7 | |
| Claude Sonnet 5 (Adaptive Reasoning, Max Effort) | 87.7 | |
| Claude Opus 4.8 (Adaptive Reasoning, Max Effort)deprecated | 87.4 | |
| Grok 4.5 (high) | 86.5 | |
| Muse Spark 1.2 (xhigh) | 82.8 | |
| Claude Opus 4.7 (Adaptive Reasoning, Max Effort)deprecated | 80.2 | |
| GPT-5.6 Sol (high) | 76.4 | |
| GLM-5.2 (max)open | 73.5 | |
| Claude Opus 4.6 (Adaptive Reasoning, Max Effort)deprecated | 73.4 | |
| MiniMax-M3open | 70.8 | |
| GPT-5.5 (high)deprecated | 68.8 | |
| GPT-5.6 Luna (xhigh) | 68.7 | |
| DeepSeek V4 Pro 0813 (Reasoning, Max Effort)open | 68.7 | |
| GPT-5.4 (xhigh)deprecated | 68.7 | |
| GPT-5.6 Terra (high) | 67.5 | |
| Sonnet 4.6 (medium) | 67.4 | |
| Kimi K2.7 Codeopen | 67.3 | |
| Kimi K2.6opendeprecated | 67.3 | |
| DeepSeek V4 Pro (Reasoning, High Effort)open | 67.2 | |
| MiMo-V2.5-Proopen | 67.2 | |
| GLM-5.1 (Reasoning)opendeprecated | 66.7 | |
| Kimi K2.5 (Reasoning)opendeprecated | 58.5 |
Pick a model on the left to start comparing. Select a second to see them head-to-head.