Claude Opus 4 wins more areas, narrowly.Grok 3 Beta is cheaper.
Scores updated · 11 tests both models report · How we compare
Where each one wins
Tests won in each of the two areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
Reasoning and images and charts rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Grok 3 Beta costs 80% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Opus 4 pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+4.2points ahead
Where Grok 3 Beta pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 11 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Other results9 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- HarmBenchClaude Opus 4 by 46.491.745.3+46.4
- AIR-Bench 2024Claude Opus 4 by 34.485.751.3+34.4
- SimpleQAGrok 3 Beta by 20.822.843.6+20.8
- Aider-PolyglotClaude Opus 4 by 17.470.753.3+17.4
- LiveCodeBenchClaude Opus 4 by 13.470.457+13.4
- MMLU-ProClaude Opus 4 by 6.786.679.9+6.7
- anthropic_red_teamClaude Opus 4 by 3.498.995.5+3.4
- simple_safety_testsClaude Opus 4 by 3.210096.8+3.2
- XSTesttie9796.4tie
Questions people ask
Which is better, Claude Opus 4 or Grok 3 Beta?
Claude Opus 4 wins one of the two areas where both have results: reasoning. Grok 3 Beta wins none, but costs 80% less. They are level on images and charts.
Which is cheaper?
Claude Opus 4 costs $15.00 per million input tokens and $75.00 per million output tokens; Grok 3 Beta costs $3.00 and $15.00. That makes Grok 3 Beta about 80% cheaper for the same work.
How do you compare the two?
We use the 11 benchmark tests both models have published scores on. The verdict counts the 2 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.