GPT-5.1 vs GPT-5.2
Wins 0 of 7 areas
—
Wins 5 of 7 areas
Coding · Agents · Reasoning · Long documents · Following instructions
GPT-5.2 is the stronger all-rounder.GPT-5.1 is cheaper.
Scores updated · 50 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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.
- 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 own12GPT-5.22 of 3 tests
- Long documentsFinding answers in very long texts01GPT-5.21 of 1 test
- Following instructionsDoing exactly what it is asked01GPT-5.21 of 1 test
- FactsGetting facts right instead of making them up22Even2 each
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 2 ties
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.
GPT-5.1 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
- Abstract visual puzzles that people can solveARC-AGI-2+35.3points ahead
- Real work tasks from 44 professionsGDPVal+31.8points ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+15points ahead
Where GPT-5.1 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+29.3points ahead
- Short factual questions, answered correctlySimpleQA Verified+10.9points ahead
- Common-sense trick questionsSimpleBench+7.4points ahead
Every test, side by side
All 50 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.2
- SciCodeGPT-5.2 by 8.843.352.1+8.8
- SWE-bench VerifiedGPT-5.2 by 3.776.380+3.7
- LMArena · WebDevGPT-5.2 by 21 rating points13951416+21 rating
- Terminal-Bench HardGPT-5.2 by 1.545.547+1.5
AgentsGPT-5.2
- GDPValGPT-5.2 by 31.816.548.3+31.8
- BrowseCompGPT-5.2 by 1550.865.8+15
- τ-Bench V3 · BankingGPT-5.1 by 4.815.911.1+4.8
ReasoningGPT-5.2
- ARC-AGI-2GPT-5.2 by 35.317.652.9+35.3
- Humanity's Last ExamGPT-5.2 by 9.228.537.7+9.2
- SimpleBenchGPT-5.1 by 7.453.245.8+7.4
- CritPtGPT-5.2 by 6.74.911.6+6.7
- GPQA DiamondGPT-5.2 by 387.390.3+3
FactsEven
- AA-Omniscience · Non-hallucinationGPT-5.1 by 29.348.118.8+29.3
- SimpleQA VerifiedGPT-5.1 by 10.94837.1+10.9
- AA-Omniscience · AccuracyGPT-5.2 by 6.637.744.3+6.6
- Vectara HHEM hallucination ratelower is betterGPT-5.2 by 1.312.110.8+1.3
Images and chartsEven
- MMMU-Protie75.574.6tie
- LMArena · Visiontie12501246tie
Other results30 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 38.621.660.2+38.6
- FrontierMath (overall)GPT-5.2 by 13.626.740.3+13.6
- ARC-AGI-1GPT-5.2 by 13.472.886.2+13.4
- matharena_visual_math_overallGPT-5.2 by 9.676.986.5+9.6
- HealthBench Hard length-adjustedGPT-5.2 by 8.925.434.3+8.9
- vectara_avg_summary_lengthGPT-5.1 by 68.1 rating points254.4186.3+68.1 rating
- Terminal-Bench 2.0GPT-5.2 by 6.447.654+6.4
- AA-OmniscienceGPT-5.1 by 6.35.4-0.9+6.3
- frontiermath_tier_4_v1GPT-5.2 by 6.312.518.8+6.3
- HealthBench Professional length-adjustedGPT-5.2 by 6.339.645.9+6.3
- AIME 2025GPT-5.2 by 694100+6
- HealthBench length-adjustedGPT-5.2 by 5.950.956.8+5.9
- AA IntelligenceGPT-5.2 by 5.724.730.4+5.7
- HMMT Nov. 2025GPT-5.2 by 4.191.795.8+4.1
- τ²-Bench (Retail)GPT-5.2 by 4.177.982+4.1
- HMMT Feb. 2025GPT-5.2 by 3.796.3100+3.7
- swe_bench_bashGPT-5.2 by 36669+3
- τ²-Bench Telecom (AA run)GPT-5.2 by 2.981.984.8+2.9
- HLE (with tools)GPT-5.2 by 2.842.745.5+2.8
- τ²-BenchGPT-5.2 by 2.882.785.5+2.8
- MMMLUGPT-5.1 by 1.49189.6+1.4
- vectara_factual_consistencyGPT-5.2 by 1.387.989.2+1.3
- Artificial Analysis Coding Indextie49.448.7tie
- HealthBench Consensus length-adjustedtie9594.4tie
- MMLU-Protie8786.7tie
- Image input eval - extremism (not_unsafe)tie11tie
- Image input eval - harms-erotic (not_unsafe)tie11tie
- Image input eval - hate (not_unsafe)tie11tie
- Image input eval - self-harm (not_unsafe)tie11tie
- vectara_answer_ratetie100100tie
Questions people ask
Which is better, GPT-5.1 or GPT-5.2?
GPT-5.2 wins five of the seven areas where both have results: coding, agents, reasoning, long documents and following instructions. GPT-5.1 wins none, but costs 29% less. They are level on facts and images and charts.
Which is better for coding?
GPT-5.2. It wins 4 of the 4 coding tests both models report; GPT-5.1 wins none.
Which is cheaper?
GPT-5.1 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 GPT-5.1 about 29% cheaper for the same work.
How do you compare the two?
We use the 50 benchmark tests both models have published scores on. The verdict counts the 20 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 30 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.