Gemini 2.5 Pro vs GPT-5.2
Wins 1 of 8 areas
Images and charts
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Following instructions
GPT-5.2 is the stronger all-rounder.Gemini 2.5 Pro is cheaper and better at images and charts.
Scores updated · 48 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software04GPT-5.24 of 4 tests
- ReasoningHard problems that need careful thinking14GPT-5.24 of 5 tests
- AgentsCarrying out multi-step tasks on its own03GPT-5.23 of 3 tests
- MathCompetition and research-level math02GPT-5.22 of 2 tests
- Following instructionsDoing exactly what it is asked01GPT-5.21 of 1 test
- Images and chartsUnderstanding pictures, charts and video41Gemini 2.5 Pro4 of 6 tests · 1 tie
- FactsGetting facts right instead of making them up22Even2 each
- Long documentsFinding answers in very long texts11Even1 each
Following instructions rests on a single test.
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 2.5 Pro costs 29% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-5.2 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+55.9points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+48points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+42.8points ahead
Where Gemini 2.5 Pro pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+18.9points ahead
- Reads and reasons about text in imagesOCRBenchv2+8.8points ahead
- Common-sense trick questionsSimpleBench+5.8points ahead
Every test, side by side
All 48 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.2
- Terminal-Bench HardGPT-5.2 by 20.526.547+20.5
- LMArena · WebDevGPT-5.2 by 192 rating points12241416+192 rating
- SWE-bench VerifiedGPT-5.2 by 16.863.280+16.8
- SciCodeGPT-5.2 by 5.846.352.1+5.8
AgentsGPT-5.2
- BrowseCompGPT-5.2 by 55.99.965.8+55.9
- GDPValGPT-5.2 by 48.3048.3+48.3
- τ-Bench V3 · BankingGPT-5.2 by 1.49.711.1+1.4
ReasoningGPT-5.2
- ARC-AGI-2GPT-5.2 by 484.952.9+48
- Humanity's Last ExamGPT-5.2 by 19.71837.7+19.7
- CritPtGPT-5.2 by 92.611.6+9
- GPQA DiamondGPT-5.2 by 6.783.690.3+6.7
- SimpleBenchGemini 2.5 Pro by 5.851.645.8+5.8
FactsEven
- SimpleQA VerifiedGemini 2.5 Pro by 18.95637.1+18.9
- AA-Omniscience · Non-hallucinationGPT-5.2 by 9.79.118.8+9.7
- AA-Omniscience · AccuracyGPT-5.2 by 5.23944.3+5.2
- Vectara HHEM hallucination ratelower is betterGemini 2.5 Pro by 3.8710.8+3.8
Images and chartsGemini 2.5 Pro
- OCRBenchv2Gemini 2.5 Pro by 8.859.350.5+8.8
- OCRBenchGemini 2.5 Pro by 5.285.980.7+5.2
- LMArena · VisionGemini 2.5 Pro by 17 rating points12631246+17 rating
- Video-MMEGPT-5.2 by 1.284.886+1.2
- MathVistaGemini 2.5 Pro by 1.183.982.8+1.1
- MMMU-Protie74.974.6tie
MathGPT-5.2
- FrontierMath Tiers 1-3 (v2)GPT-5.2 by 42.824.667.4+42.8
- FrontierMath Tier 4GPT-5.2 by 31.7031.7+31.7
Long documentsEven
- AA-LCRGPT-5.2 by 13.76982.7+13.7
- LongBench v2Gemini 2.5 Pro by 4.358.854.5+4.3
Other results21 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.
- AA Agentic IndexGPT-5.2 by 56.73.560.2+56.7
- ARC-AGI-1GPT-5.2 by 49.23786.2+49.2
- HMMT 2025GPT-5.2 by 33.765.799.4+33.7
- τ²-BenchGPT-5.2 by 31.55485.5+31.5
- τ²-Bench Telecom (AA run)GPT-5.2 by 30.754.184.8+30.7
- HMMT Feb. 2025GPT-5.2 by 19.280.8100+19.2
- HLE (with tools)GPT-5.2 by 17.128.445.5+17.1
- AIME 2025GPT-5.2 by 1783100+17
- HMMT Nov. 2025GPT-5.2 by 15.88095.8+15.8
- AA IntelligenceGPT-5.2 by 15.41530.4+15.4
- AA-OmniscienceGPT-5.2 by 15.4-16.4-0.9+15.4
- LiveCodeBenchGPT-5.2 by 14.874.289+14.8
- frontiermath_tier_4_v1GPT-5.2 by 14.64.218.8+14.6
- MATH-VisionGPT-5.2 by 9.773.383+9.7
- matharena_visual_math_overallGPT-5.2 by 9.377.286.5+9.3
- vectara_avg_summary_lengthGPT-5.2 by 79.9 rating points106.4186.3+79.9 rating
- vectara_factual_consistencyGemini 2.5 Pro by 3.89389.2+3.8
- VideoMMMUGPT-5.2 by 2.383.685.9+2.3
- Artificial Analysis Coding IndexGPT-5.2 by 246.748.7+2
- vectara_answer_ratetie99.1100tie
- MMLU-Protie8686.7tie
Questions people ask
Which is better, Gemini 2.5 Pro or GPT-5.2?
GPT-5.2 wins five of the eight areas we test: coding, agents, reasoning, math and following instructions. Gemini 2.5 Pro wins images and charts, and costs 29% less. They are level on facts and long documents.
Which is better for coding?
GPT-5.2. It wins 4 of the 4 coding tests both models report; Gemini 2.5 Pro wins none.
Which is cheaper?
Gemini 2.5 Pro costs $1.25 per million input tokens and $10.00 per million output tokens; GPT-5.2 costs $1.75 and $14.00. That makes Gemini 2.5 Pro about 29% cheaper for the same work.
How do you compare the two?
We use the 48 benchmark tests both models have published scores on. The verdict counts the 27 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 21 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.