GPT-4.1 vs o4-mini
Wins 2 of 8 areas
Facts · Long documents
Wins 6 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math · Following instructions
o4-mini is the stronger all-rounder.GPT-4.1 is better at facts.
Scores updated · 41 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.
- Images and chartsUnderstanding pictures, charts and video14o4-mini4 of 5 tests
- CodingWriting and fixing software03o4-mini3 of 3 tests
- ReasoningHard problems that need careful thinking02o4-mini2 of 3 tests · 1 tie
- Following instructionsDoing exactly what it is asked02o4-mini2 of 2 tests
- AgentsCarrying out multi-step tasks on its own01o4-mini1 of 1 test
- MathCompetition and research-level math01o4-mini1 of 1 test
- FactsGetting facts right instead of making them up31GPT-4.13 of 4 tests
- Long documentsFinding answers in very long texts10GPT-4.11 of 1 test
Agents, math and long documents 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.
o4-mini costs 45% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where o4-mini pulls ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+30.1points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+25.7points ahead
- Reasoning about charts from research papersCharXiv (RQ)+15.3points ahead
Where GPT-4.1 pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+12.3points ahead
- Reasons across sets of long documentsAA-LCR+7.3points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+3points ahead
Every test, side by side
All 41 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Codingo4-mini
- SWE-bench Verifiedo4-mini by 13.554.668.1+13.5
- SciCodeo4-mini by 8.438.146.5+8.4
- Terminal-Bench Hardo4-mini by 1.613.615.2+1.6
Reasoningo4-mini
- Humanity's Last Examo4-mini by 12.34.216.5+12.3
- GPQA Diamondo4-mini by 11.866.678.4+11.8
- CritPttie00.6tie
FactsGPT-4.1
- Vectara HHEM hallucination ratelower is betterGPT-4.1 by 135.618.6+13
- AA-Omniscience · Non-hallucinationo4-mini by 12.86.719.5+12.8
- SimpleQA VerifiedGPT-4.1 by 12.331.118.8+12.3
- AA-Omniscience · AccuracyGPT-4.1 by 327.824.8+3
Images and chartso4-mini
- CharXiv (RQ)o4-mini by 15.356.772+15.3
- MathVistao4-mini by 12.172.284.3+12.1
- MMMU-Proo4-mini by 861.269.2+8
- MMMUo4-mini by 6.874.881.6+6.8
- LMArena · VisionGPT-4.1 by 16 rating points12101194+16 rating
Following instructionso4-mini
- IFBencho4-mini by 25.74368.7+25.7
- Multi-Challengeo4-mini by 4.738.343+4.7
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.
- ARC-AGI-1o4-mini by 53.25.558.7+53.2
- AIME 2025o4-mini by 46.346.492.7+46.3
- Aider-Polygloto4-mini by 17.351.668.9+17.3
- AIR-Bench 2024o4-mini by 13.764.878.5+13.7
- vectara_factual_consistencyGPT-4.1 by 1394.481.4+13
- AA Agentic Indexo4-mini by 8.827.336.1+8.8
- τ²-Bench Telecom (AA run)o4-mini by 8.547.155.6+8.5
- frontiermath_tier_4_v1o4-mini by 6.306.3+6.3
- HarmBencho4-mini by 5.391.797+5.3
- AA Intelligenceo4-mini by 412.716.7+4
- AA-Omniscienceo4-mini by 3.9-39.6-35.7+3.9
- Artificial Analysis Coding Indexo4-mini by 3.821.825.6+3.8
- TAU-bench (retail)o4-mini by 3.86871.8+3.8
- vectara_avg_summary_lengtho4-mini by 36 rating points91.7127.7+36 rating
- MMLU-ProXGPT-4.1 by 3.472.769.3+3.4
- bbqo4-mini by 1.492.694+1.4
- anthropic_red_teamGPT-4.1 by 1.199.398.2+1.1
- vectara_answer_ratetie99.999.2tie
- XSTesttie97.997.4tie
- TAU-bench (airline)tie49.449.2tie
- simple_safety_teststie100100tie
Questions people ask
Which is better, GPT-4.1 or o4-mini?
o4-mini wins six of the eight areas we test: coding, agents, reasoning, images and charts, math and following instructions. GPT-4.1 wins facts and long documents.
Which is better for coding?
o4-mini. It wins 3 of the 3 coding tests both models report; GPT-4.1 wins none.
Which is cheaper?
GPT-4.1 costs $2.00 per million input tokens and $8.00 per million output tokens; o4-mini costs $1.10 and $4.40. That makes o4-mini about 45% cheaper for the same work.
How do you compare the two?
We use the 41 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 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.