Grok vs Claude
As of November 2026, Claude is ahead of Grok on public benchmark score: Claude Opus 5.5 scores 86.32 at #2, against Grok 4.6 at 66.76 and #23. Grok's top-ranked model is the cheaper one: Grok 4.6 at $2/$6 per million input/output tokens, against Claude Opus 5.5 at $4/$20.
This page compares Grok and Claude as products, through the models behind each one: xAI's Grok 4.7, Grok 4.6, and Grok 4.1 Fast and Anthropic's Claude Opus 5.5, Claude Fable 5.1, Claude Sonnet 5.5, and Claude Haiku 5.5. Every number below is our sourced benchmark, preference, or pricing data for those specific models.
Search it as "Grok vs Claude" or "Claude vs Grok": the score column follows the same public ranks as the main leaderboard, while prices and raw benchmark rows stay separate evidence.
The short answer
The headline number favors Claude Opus 5.5 at 86.32 (#2). Grok's highest-ranked row here is Grok 4.6 at 66.76 (#23). Grok 4.7, xAI's newest flagship, is not publicly ranked, so this comparison uses the highest-ranked Grok above. Its published results are on its model page.
On coding, Claude Opus 5.5 ranks #1 against #22 for Grok 4.6. On agentic work, Claude Opus 5.5 ranks #1 against #10 for Grok 4.6. On blind human preference, Claude Opus 5.5 (1508.7) is clear of Grok 4.6 (1455.5), the best-rated model on the other side.
Then there is price. If you want the highest benchmarked ceiling today, it is Claude; if you want the cheaper top-ranked model, it is Grok. At the budget end the order flips: Claude Haiku 5.5 lists at $0.10 / $0.50, against $0.20 / $0.50 for Grok 4.1 Fast.
Grok vs Claude: the models behind the apps
Score and coding are the public overall and coding lanes, with rank in brackets. The preference rating is an independent blind human-preference vote, with its ± margin; two ratings whose ranges overlap are a tie.
| Model | Score | Coding | Preference rating | Context |
|---|---|---|---|---|
| Grok 4.7xAI | Not publicly ranked | — | 1441.2 ±5.7 | 500K |
| Grok 4.6xAI | 66.76 (#23) | 58.57 (#22) | 1455.5 ±4.6 | 500K |
| Grok 4.1 FastxAI | 36.39 (#154) | — | 1420.0 | 1M |
| Claude Opus 5.5Anthropic | 86.32 (#2) | 82.82 (#1) | 1508.7 ±6.6 | 1M |
| Claude Fable 5.1Anthropic | 81.3 (#7) | 76.58 (#4) | 1504.4 ±5.8 | 1M |
| Claude Sonnet 5.5Anthropic | 81.78 (#6) | 81.88 (#2) | 1477.2 ±6.8 | 1M |
| Claude Haiku 5.5Anthropic | 66.43 (#25) | 52.52 (#41) | 1424.8 ±7.5 | 1M |
For the benchmark-by-benchmark detail, the model comparison pages go deeper: Grok 4.6 vs Claude Opus 5.5 is the top-ranked matchup, and Grok 4.1 Fast vs Claude Haiku 5.5 covers the budget tiers.
Grok vs Claude API pricing
The API is where the two compare exactly. Claude Haiku 5.5 is the cheapest model on this page at $0.10 / $0.50 per million input / output tokens, and Claude Fable 5.1 the most expensive at $10 / $50. The full Claude rate card, including cached input and batch rates, is on the Anthropic API pricing hub, and every Grok rate is on each model's page. Consumer plans differ more than token rates: compare the usage limits of each plan you are considering.
| Model | Input / 1M tokens | Output / 1M tokens |
|---|---|---|
| Grok 4.7 | $2 | $6 |
| Grok 4.6 | $2 | $6 |
| Grok 4.1 Fast | $0.20 | $0.50 |
| Claude Opus 5.5 | $4 | $20 |
| Claude Fable 5.1 | $10 | $50 |
| Claude Sonnet 5.5 | $2 | $10 |
| Claude Haiku 5.5 | $0.10 | $0.50 |
Which should you pick?
Pick Grok if…
- Price at the top matters most: Grok 4.6 costs $2 / $6 per million tokens against $4 / $20 for Claude Opus 5.5.
- You want xAI's newest model, Grok 4.7, which is not publicly ranked. Test it on your own tasks first.
Pick Claude if…
- You want the higher benchmarked ceiling: Claude Opus 5.5 ranks #2 overall.
- Coding is the job: Claude Opus 5.5 ranks #1 on the public coding ranking, against #22 for Grok 4.6.
- Agents and long tool loops count: Claude Opus 5.5 ranks #1 on the public agentic ranking.
- Writing quality matters: Claude Opus 5.5 leads on blind human preference.
- You run high volume on a budget model: Claude Haiku 5.5 lists at $0.10 / $0.50.
How this comparison works
We compare products through the models that power them, because both xAI and Anthropic ship new models several times a year and the rows here follow each release. Overall, coding, and agentic scores use the current public ranking lanes; a model with only provider-reported results is shown as not publicly ranked. Preference ratings come from an independent blind human-preference index, and API prices come from each provider's rate card as recorded in our pricing data. The model comparison pages use the same sourced dataset.
Questions
Is Grok better than Claude?
No, not on the current public score: Claude Opus 5.5 posts 86.32 at #2, while Grok 4.6 posts 66.76 at #23. Grok 4.7 is not publicly ranked.
Is Grok cheaper than Claude?
At the top, yes; at the budget end, no. Grok 4.6 costs $2 input / $6 output per million tokens against $4 / $20 for Claude Opus 5.5, each side's top-ranked model. The cheapest model on each side is Grok 4.1 Fast at $0.20 / $0.50 and Claude Haiku 5.5 at $0.10 / $0.50.
Which is better for coding, Grok or Claude?
Claude, on the public coding ranking: Claude Opus 5.5 ranks #1 at 82.82, against Grok 4.6 at #22 with 58.57. Claude Opus 5.5 leads by point estimate; its evidence label is Supported. The conditional ranges do not establish rank confidence.
Which writes better, Grok or Claude?
Claude, on blind votes: Claude Opus 5.5 rates 1520.2 ±13.1 on the independent blind creative-writing vote, against 1441.8 ±8.0 for Grok 4.6, the best-rated writer on the other side. The ranges do not overlap, so the gap is more than noise.