Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Claude 4 Sonnet is clearly ahead on the aggregate, 38 to 23. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Claude 4 Sonnet's sharpest advantage is in reasoning, where it averages 69.9 against 21.7. The single biggest benchmark swing on the page is SimpleQA, 71% to 21.7%. Grok 3 Mini does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.
Claude 4 Sonnet gives you the larger context window at 200K, compared with 128K for Grok 3 Mini.
Pick Claude 4 Sonnet if you want the stronger benchmark profile. Grok 3 Mini only becomes the better choice if knowledge is the priority.
Benchmark data for this category is coming soon.
Claude 4 Sonnet
68.9
Grok 3 Mini
41.5
Benchmark data for this category is coming soon.
Claude 4 Sonnet
69.9
Grok 3 Mini
21.7
Claude 4 Sonnet
70.5
Grok 3 Mini
74.1
Benchmark data for this category is coming soon.
Benchmark data for this category is coming soon.
Claude 4 Sonnet
72
Grok 3 Mini
39.7
Claude 4 Sonnet is ahead overall, 38 to 23. The biggest single separator in this matchup is SimpleQA, where the scores are 71% and 21.7%.
Grok 3 Mini has the edge for knowledge tasks in this comparison, averaging 74.1 versus 70.5. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Claude 4 Sonnet has the edge for coding in this comparison, averaging 68.9 versus 41.5. Grok 3 Mini stays close enough that the answer can still flip depending on your workload.
Claude 4 Sonnet has the edge for math in this comparison, averaging 72 versus 39.7. Inside this category, AIME 2024 is the benchmark that creates the most daylight between them.
Claude 4 Sonnet has the edge for reasoning in this comparison, averaging 69.9 versus 21.7. Inside this category, SimpleQA is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.