GPT-5.4 nano vs GPT-5.5
Wins 0 of 8 areas
—
Wins 7 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math · Long documents · Following instructions
GPT-5.5 is the stronger all-rounder.GPT-5.4 nano is cheaper.
Scores updated · 46 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.
- AgentsCarrying out multi-step tasks on its own07GPT-5.57 of 7 tests
- CodingWriting and fixing software06GPT-5.56 of 6 tests
- ReasoningHard problems that need careful thinking05GPT-5.55 of 5 tests
- MathCompetition and research-level math03GPT-5.53 of 3 tests
- Images and chartsUnderstanding pictures, charts and video02GPT-5.52 of 2 tests
- Following instructionsDoing exactly what it is asked01GPT-5.51 of 2 tests · 1 tie
- Long documentsFinding answers in very long texts01GPT-5.51 of 1 test
- FactsGetting facts right instead of making them up22Even2 each
Long documents 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.
GPT-5.4 nano costs 96% 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-5.5 pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+79.3points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+60.3points ahead
- Short factual questions, answered correctlySimpleQA Verified+51.3points ahead
Where GPT-5.4 nano pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+14.8points ahead
Every test, side by side
All 46 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.5
- Terminal-Bench 2.1GPT-5.5 by 23.660.784.3+23.6
- Terminal-Bench HardGPT-5.5 by 18.242.460.6+18.2
- LiveBench · CodingGPT-5.5 by 11.470.882.2+11.4
- SciCodeGPT-5.5 by 8.647.255.8+8.6
- LiveBench · Agentic CodingGPT-5.5 by 7.246.854+7.2
- SWE-bench ProGPT-5.5 by 6.252.458.6+6.2
AgentsGPT-5.5
- OSWorld-VerifiedGPT-5.5 by 39.73978.7+39.7
- AA IT-Bench SREGPT-5.5 by 21.424.445.8+21.4
- GDPValGPT-5.5 by 20.222.542.7+20.2
- MCP AtlasGPT-5.5 by 19.256.175.3+19.2
- Terminal-Bench 4.0GPT-5.5 by 14.10.514.6+14.1
- AA ApexAgentsGPT-5.5 by 12.824.937.7+12.8
- τ-Bench V3 · BankingGPT-5.5 by 11.627.439+11.6
ReasoningGPT-5.5
- ARC-AGI-2GPT-5.5 by 79.35.785+79.3
- CritPtGPT-5.5 by 17.89.327.1+17.8
- Humanity's Last ExamGPT-5.5 by 17.528.345.8+17.5
- GPQA DiamondGPT-5.5 by 11.881.793.5+11.8
- LiveBench · ReasoningGPT-5.5 by 8.681.189.7+8.6
FactsEven
- SimpleQA VerifiedGPT-5.5 by 51.311.763+51.3
- AA-Omniscience · AccuracyGPT-5.5 by 32.325.758+32.3
- AA-Omniscience · Non-hallucinationGPT-5.4 nano by 14.825.811+14.8
- Vectara HHEM hallucination ratelower is betterGPT-5.4 nano by 6.23.19.3+6.2
Images and chartsGPT-5.5
- MMMU-ProGPT-5.5 by 14.565.479.9+14.5
- LMArena · VisionGPT-5.5 by 96 rating points11961292+96 rating
MathGPT-5.5
- FrontierMath Tier 4GPT-5.5 by 60.312.272.5+60.3
- FrontierMath Tiers 1-3 (v2)GPT-5.5 by 40.444.985.3+40.4
- LiveBench · MathematicsGPT-5.5 by 4.99195.9+4.9
Following instructionsGPT-5.5
- LiveBench · Instruction FollowingGPT-5.5 by 3.567.270.7+3.5
- IFBenchtie75.975.9tie
Other results16 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-OmniscienceGPT-5.5 by 50-29.520.5+50
- ARC-AGI-1GPT-5.5 by 43.551.595+43.5
- Terminal-Bench 2.0GPT-5.5 by 36.446.382.7+36.4
- frontiermath_tier_4_v1GPT-5.5 by 29.16.335.4+29.1
- livebench_languageGPT-5.5 by 24.962.587.4+24.9
- ToolathlonGPT-5.5 by 20.135.555.6+20.1
- AA Agentic IndexGPT-5.5 by 19.617.737.3+19.6
- Artificial Analysis Coding IndexGPT-5.5 by 18.856.174.9+18.8
- τ²-Bench Telecom (AA run)GPT-5.5 by 17.97693.9+17.9
- AA IntelligenceGPT-5.5 by 17.720.738.4+17.7
- livebench_data_analysisGPT-5.5 by 1467.681.6+14
- Finance Agent v2GPT-5.5 by 13.638.251.8+13.6
- LiveBenchGPT-5.5 by 10.670.180.7+10.6
- vectara_factual_consistencyGPT-5.4 nano by 6.296.990.7+6.2
- vectara_avg_summary_lengthGPT-5.4 nano by 14.8 rating points144.4129.6+14.8 rating
- vectara_answer_ratetie100100tie
Questions people ask
Which is better, GPT-5.4 nano or GPT-5.5?
GPT-5.5 wins seven of the eight areas we test: coding, agents, reasoning, images and charts, math, long documents and following instructions. GPT-5.4 nano wins none, but costs 96% less. They are level on facts.
Which is better for coding?
GPT-5.5. It wins 6 of the 6 coding tests both models report; GPT-5.4 nano wins none.
Which is cheaper?
GPT-5.4 nano costs $0.20 per million input tokens and $1.25 per million output tokens; GPT-5.5 costs $5.00 and $30.00. That makes GPT-5.4 nano about 96% cheaper for the same work.
How do you compare the two?
We use the 46 benchmark tests both models have published scores on. The verdict counts the 30 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 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.