Gemini 2.5 Flash vs Grok 3 Mini Reasoning
Wins 4 of 6 areas
Agents · Reasoning · Long documents · Following instructions
Wins 1 of 6 areas
Coding
Gemini 2.5 Flash is the stronger all-rounder.Grok 3 Mini Reasoning is cheaper and better at coding.
Scores updated · 21 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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.
- ReasoningHard problems that need careful thinking10Gemini 2.5 Flash1 of 3 tests · 2 ties
- AgentsCarrying out multi-step tasks on its own10Gemini 2.5 Flash1 of 1 test
- Long documentsFinding answers in very long texts10Gemini 2.5 Flash1 of 1 test
- Following instructionsDoing exactly what it is asked10Gemini 2.5 Flash1 of 1 test
- CodingWriting and fixing software02Grok 3 Mini Reasoning2 of 2 tests
- FactsGetting facts right instead of making them up11Even1 each
Agents, long documents and following instructions 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 Mini Reasoning costs 71% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemini 2.5 Flash pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+10.9points ahead
- Reasons across sets of long documentsAA-LCR+10points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+4.4points ahead
Where Grok 3 Mini Reasoning pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+49.5points ahead
- Code for real scientific research problemsSciCode+11.5points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+3.8points ahead
Every test, side by side
All 21 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGrok 3 Mini Reasoning
- SciCodeGrok 3 Mini Reasoning by 11.529.140.6+11.5
- Terminal-Bench HardGrok 3 Mini Reasoning by 3.813.617.4+3.8
ReasoningGemini 2.5 Flash
- Humanity's Last ExamGemini 2.5 Flash by 1.112.111+1.1
- CritPttie1.10.6tie
- GPQA Diamondtie7979.1tie
FactsEven
- AA-Omniscience · Non-hallucinationGrok 3 Mini Reasoning by 49.524.774.2+49.5
- AA-Omniscience · AccuracyGemini 2.5 Flash by 10.925.915.1+10.9
Following instructionsGemini 2.5 Flash
- IFBenchGemini 2.5 Flash by 4.450.345.9+4.4
Other results11 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.
- τ²-Bench Telecom (AA run)Grok 3 Mini Reasoning by 58.831.690.4+58.8
- AA-OmniscienceGrok 3 Mini Reasoning by 23-29.8-6.8+23
- AIME 2025Grok 3 Mini Reasoning by 18.87290.8+18.8
- ARC-AGI-1Gemini 2.5 Flash by 15.832.316.5+15.8
- AA Agentic IndexGrok 3 Mini Reasoning by 12.418.831.2+12.4
- LiveCodeBenchGrok 3 Mini Reasoning by 4.276.280.4+4.2
- livecodebench_mediumGrok 3 Mini Reasoning by 3.482.585.9+3.4
- Artificial Analysis Coding IndexGrok 3 Mini Reasoning by 322.225.2+3
- livecodebench_hardGrok 3 Mini Reasoning by 2.848.351.1+2.8
- AA IntelligenceGrok 3 Mini Reasoning by 1.513.114.6+1.5
- livecodebench_easyGemini 2.5 Flash by 199.198.1+1
Questions people ask
Which is better, Gemini 2.5 Flash or Grok 3 Mini Reasoning?
Gemini 2.5 Flash wins four of the six areas where both have results: agents, reasoning, long documents and following instructions. Grok 3 Mini Reasoning wins coding, and costs 71% less. They are level on facts.
Which is better for coding?
Grok 3 Mini Reasoning. It wins 2 of the 2 coding tests both models report; Gemini 2.5 Flash wins none.
Which is cheaper?
Gemini 2.5 Flash costs $0.30 per million input tokens and $2.50 per million output tokens; Grok 3 Mini Reasoning costs $0.30 and $0.50. That makes Grok 3 Mini Reasoning about 71% cheaper for the same work.
How do you compare the two?
We use the 21 benchmark tests both models have published scores on. The verdict counts the 10 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 11 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.