Gemini 3 Pro vs Grok 3
Wins 5 of 6 areas
Coding · Agents · Reasoning · Long documents · Following instructions
Wins 1 of 6 areas
Facts
Gemini 3 Pro is the stronger all-rounder.Grok 3 is better at facts.
Scores updated · 26 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 thinking50Gemini 3 Pro5 of 5 tests
- CodingWriting and fixing software20Gemini 3 Pro2 of 2 tests
- AgentsCarrying out multi-step tasks on its own10Gemini 3 Pro1 of 1 test
- Long documentsFinding answers in very long texts10Gemini 3 Pro1 of 1 test
- Following instructionsDoing exactly what it is asked10Gemini 3 Pro1 of 1 test
- FactsGetting facts right instead of making them up12Grok 32 of 3 tests
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.
Gemini 3 Pro costs 42% 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 Gemini 3 Pro pulls ahead
- Common-sense trick questionsSimpleBench+40.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+35.6points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+30.3points ahead
Where Grok 3 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+4.2points ahead
Every test, side by side
All 26 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3 Pro
- Terminal-Bench HardGemini 3 Pro by 30.341.711.4+30.3
- SciCodeGemini 3 Pro by 19.356.136.8+19.3
ReasoningGemini 3 Pro
- SimpleBenchGemini 3 Pro by 40.376.436.1+40.3
- Humanity's Last ExamGemini 3 Pro by 35.639.74.1+35.6
- ARC-AGI-2Gemini 3 Pro by 31.131.10+31.1
- GPQA DiamondGemini 3 Pro by 21.590.869.3+21.5
- CritPtGemini 3 Pro by 9.19.10+9.1
FactsGrok 3
- AA-Omniscience · AccuracyGemini 3 Pro by 26.955.828.9+26.9
- Vectara HHEM hallucination ratelower is betterGrok 3 by 7.813.65.8+7.8
- AA-Omniscience · Non-hallucinationGrok 3 by 4.28.512.7+4.2
Following instructionsGemini 3 Pro
- IFBenchGemini 3 Pro by 23.570.446.9+23.5
Other results13 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.
- ARC-AGI-1Gemini 3 Pro by 69.5755.5+69.5
- AA-OmniscienceGemini 3 Pro by 48.515.3-33.2+48.5
- τ²-Bench Telecom (AA run)Gemini 3 Pro by 38.387.148.8+38.3
- AA Agentic IndexGemini 3 Pro by 27.45224.6+27.4
- Artificial Analysis Coding IndexGemini 3 Pro by 26.746.519.8+26.7
- frontiermath_tier_4_v1Gemini 3 Pro by 18.818.80+18.8
- AA IntelligenceGemini 3 Pro by 15.92812.1+15.9
- LiveCodeBenchGemini 3 Pro by 12.69279.4+12.6
- SimpleQAGrok 3 by 9.972.182+9.9
- vectara_factual_consistencyGrok 3 by 7.886.494.2+7.8
- AIME 2025Gemini 3 Pro by 6.710093.3+6.7
- vectara_answer_rateGemini 3 Pro by 6.499.493+6.4
- vectara_avg_summary_lengthtie101.995.9tie
Questions people ask
Which is better, Gemini 3 Pro or Grok 3?
Gemini 3 Pro wins five of the six areas where both have results: coding, agents, reasoning, long documents and following instructions. Grok 3 wins facts.
Which is better for coding?
Gemini 3 Pro. It wins 2 of the 2 coding tests both models report; Grok 3 wins none.
Which is cheaper?
Gemini 3 Pro costs $2.00 per million input tokens and $12.00 per million output tokens; Grok 3 costs $4.00 and $20.00. That makes Gemini 3 Pro about 42% cheaper for the same work.
How do you compare the two?
We use the 26 benchmark tests both models have published scores on. The verdict counts the 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 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.