Head-to-head · 4 models compared
Claude vs Grok
Anthropic and xAI between them account for 4 rated models on the OpenRating Index. Claude's strongest entry is Claude Opus 5 (max) at 63.4, against Grok's Grok 4.6 (high) at 60.8. Both are scored on the same axes, with the same tests, below.
Who wins on each axis
Claude leads all three axes in this comparison.
OpenRating Index
Claude
63.4
Claude: Claude Opus 5 (max) · composite 0–100
Output speed
Claude
195 tok/s
Claude: Claude 4.5 Haiku · blended tok/s, fastest model
Cost / task
Claude
$0.22
Claude: Claude 4.5 Haiku · lowest model, per benchmark run
Every rated model, side by side
Sorted by OpenRating Index. Every figure is measured under the same standardised workload.
| Model | Provider | Index | Reasoning | Speed | Cost / task | Input / 1M |
|---|---|---|---|---|---|---|
| Claude Opus 5 (max) | Claude | 63.4 | 89.4 | 59 tok/s | $2.34 | $4.8 |
| Claude Fable 5 (with fallback) | Claude | 62.3 | 87.8 | 72 tok/s | $3.14 | $5.5 |
| Grok 4.6 (high) | Grok | 60.8 | 83.6 | 69 tok/s | $0.84 | $1.8 |
| Claude 4.5 Haiku | Claude | 49.6 | 68.2 | 195 tok/s | $0.22 | $0.4 |
Put Claude and Grok head to head yourself
Pick any combination of models and compare their full cost breakdown — reasoning, answer generation, cache overhead, and input payload — in the OpenRating comparison tool.
Open the comparison tool →