Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Grok 3 Mini is clearly ahead on the aggregate, 23 to 19. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Grok 3 Mini's sharpest advantage is in knowledge, where it averages 74.1 against 40.2. The single biggest benchmark swing on the page is AIME 2024, 8.6% to 39.7%. GPT-4o mini does hit back in coding, so the answer changes if that is the part of the workload you care about most.
GPT-4o mini is also the more expensive model on tokens at $0.15 input / $0.60 output per 1M tokens, versus $0.30 input / $0.50 output per 1M tokens for Grok 3 Mini.
Pick Grok 3 Mini if you want the stronger benchmark profile. GPT-4o mini only becomes the better choice if coding is the priority.
Benchmark data for this category is coming soon.
GPT-4o mini
48
Grok 3 Mini
41.5
Benchmark data for this category is coming soon.
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
GPT-4o mini
40.2
Grok 3 Mini
74.1
Comparable scores for this category are coming soon. One or both models do not have sourced results here yet.
Benchmark data for this category is coming soon.
GPT-4o mini
8.6
Grok 3 Mini
39.7
Grok 3 Mini is ahead overall, 23 to 19. The biggest single separator in this matchup is AIME 2024, where the scores are 8.6% and 39.7%.
Grok 3 Mini has the edge for knowledge tasks in this comparison, averaging 74.1 versus 40.2. Inside this category, GPQA is the benchmark that creates the most daylight between them.
GPT-4o mini has the edge for coding in this comparison, averaging 48 versus 41.5. Grok 3 Mini stays close enough that the answer can still flip depending on your workload.
Grok 3 Mini has the edge for math in this comparison, averaging 39.7 versus 8.6. Inside this category, AIME 2024 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.