GPT-4o vs GPT-5.4
Wins 0 of 7 areas
—
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Images and charts · Long documents · Following instructions
GPT-5.4 is the stronger all-rounder.
Scores updated · 29 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.
- ReasoningHard problems that need careful thinking04GPT-5.44 of 4 tests
- FactsGetting facts right instead of making them up13GPT-5.43 of 4 tests
- CodingWriting and fixing software02GPT-5.42 of 2 tests
- AgentsCarrying out multi-step tasks on its own02GPT-5.42 of 2 tests
- Images and chartsUnderstanding pictures, charts and video02GPT-5.42 of 2 tests
- Long documentsFinding answers in very long texts01GPT-5.41 of 1 test
- Following instructionsDoing exactly what it is asked01GPT-5.41 of 1 test
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-5.4 costs 13% 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 GPT-5.4 pulls ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+70points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+49.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+41.9points ahead
Where GPT-4o pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+53.8points ahead
Every test, side by side
All 29 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.4
- Terminal-Bench HardGPT-5.4 by 49.38.357.6+49.3
- SciCodeGPT-5.4 by 23.333.356.6+23.3
ReasoningGPT-5.4
- ARC-AGI-2GPT-5.4 by 73.3073.3+73.3
- Humanity's Last ExamGPT-5.4 by 41.91.843.7+41.9
- GPQA DiamondGPT-5.4 by 39.452.692+39.4
- CritPtGPT-5.4 by 23.4023.4+23.4
FactsGPT-5.4
- AA-Omniscience · Non-hallucinationGPT-4o by 53.862.18.3+53.8
- AA-Omniscience · AccuracyGPT-5.4 by 27.223.750.9+27.2
- SimpleQA VerifiedGPT-5.4 by 19.12645.1+19.1
- Vectara HHEM hallucination ratelower is betterGPT-5.4 by 2.69.67+2.6
Images and chartsGPT-5.4
- CharXiv (RQ)GPT-5.4 by 2458.882.8+24
- MMMU-ProGPT-5.4 by 22.156.378.4+22.1
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.
- Codeforces (Rating)GPT-5.4 by 2409 rating points7593168+2409 rating
- ARC-AGI-1GPT-5.4 by 89.24.593.7+89.2
- MATH-VisionGPT-5.4 by 61.630.492+61.6
- τ²-Bench Telecom (AA run)GPT-5.4 by 58.228.987.1+58.2
- Artificial Analysis Coding IndexGPT-5.4 by 46.924.271.1+46.9
- AA Agentic IndexGPT-5.4 by 35.88.444.2+35.8
- AA IntelligenceGPT-5.4 by 31.77.339+31.7
- Chinese SimpleQA (C-SimpleQA)GPT-5.4 by 18.158.776.8+18.1
- AA-OmniscienceGPT-5.4 by 16.3-10.55.8+16.3
- MMLU-ProGPT-5.4 by 12.874.787.5+12.8
- vectara_answer_rateGPT-5.4 by 6.193.899.9+6.1
- vectara_avg_summary_lengthGPT-4o by 4.986.681.7+4.9
- vectara_factual_consistencyGPT-5.4 by 2.690.493+2.6
Questions people ask
Which is better, GPT-4o or GPT-5.4?
GPT-5.4 wins all seven areas where both have results: coding, agents, reasoning, facts, images and charts, long documents and following instructions. GPT-4o wins none.
Which is better for coding?
GPT-5.4. It wins 2 of the 2 coding tests both models report; GPT-4o wins none.
Which is cheaper?
GPT-4o costs $5.00 per million input tokens and $15.00 per million output tokens; GPT-5.4 costs $2.50 and $15.00. That makes GPT-5.4 about 13% cheaper for the same work.
How do you compare the two?
We use the 29 benchmark tests both models have published scores on. The verdict counts the 16 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.