OpenAIreleased 23 Apr 2026

GPT-5.5

#17 of 9537 of 46 benchmarksConsistent

69.7

B.O.B.

Category profile

Real use (hard to game)2×67
Reasoning & knowledge78
Math56
Coding71
Agents & real work67
Independent indexes83

Reality check

Lab tests

Real use (8 tests)

Lab tests and real-use tests agree. Its lab-test and real-use scores are within 2 points of each other.

What counts as real use

Price per 1M tokens (OpenRouter)

$5.00 in · $30.00 out

Cost per task on the AA index

—

Output speed (Artificial Analysis)

—

§Receipts

Every benchmark result

“Beats” = the share of tracked models this one beat on that test. The raw score is its best reported setting.

Benchmark	Raw	Beats	Source
Real use (hard to game) · Real use 67
Kagi LLM Benchmarkas gpt-5-5	88.8%	93%of 22	Kagi
SWE-rebenchas gpt-5.5-2026-04-23-xhigh	62.7%	89%of 34	Nebius
LMArena Textas gpt-5.5-high	1481	73%of 80	LMArena (blind human votes)
LMArena Agent Arenaas GPT 5.5 (xHigh)	0.035	69%of 50	LMArena (real agent sessions)
EQ-Bench Creative Writingas gpt-5.5	1844	69%of 46	Sam Paech (open source)
SimpleBenchas gpt-5.5_unknown	69.0%	62%of 30	Epoch AI Benchmarking Hub
LMArena WebDevas gpt-5.5-xhigh (codex-harness)	1513	44%of 82	LMArena (blind human votes)
LiveBench 2.0 (real-work suites)as gpt-5.5	61.9%	34%of 39	LiveBench / Abacus.AI
Reasoning & knowledge · Reasoning 78
Chess Puzzlesas gpt-5.5_xhigh	63.0%	94%of 73	Epoch AI Benchmarking Hub
Mystery Game Puzzlesas gpt-5.5_xhigh	56.0%	85%of 57	Epoch AI Benchmarking Hub
CritPt (physics research)as gpt-5.5_xhigh	27.1%	85%of 89	Epoch AI Benchmarking Hub
ARC-AGI-2as gpt-5.5_xhigh	85.0%	79%of 44	Epoch AI Benchmarking Hub
SimpleQA Verifiedas gpt-5.5_xhigh	63.0%	79%of 57	Epoch AI Benchmarking Hub
Blueprint-Bench 2as gpt-5.5_unknown	36.2%	73%of 31	Epoch AI Benchmarking Hub
ForecastBenchas gpt-5.5_unknown	60.8%	71%of 18	Epoch AI Benchmarking Hub
GPQA Diamondas gpt-5.5_low	90.7%	58%of 70	Epoch AI Benchmarking Hub
Math · Math 56
FrontierMath Tiers 1-3as gpt-5.5_xhigh	85.3%	80%of 60	Epoch AI Benchmarking Hub
FrontierMath Tier 4as gpt-5.5_xhigh	72.5%	75%of 52	Epoch AI Benchmarking Hub
ProofBenchas gpt-5.5_xhigh	50.0%	55%of 68	Epoch AI Benchmarking Hub
OTIS Mock AIMEas gpt-5.5_low	84.4%	15%of 73	Epoch AI Benchmarking Hub
Coding · Coding 71
Terminal-Benchas gpt-5.5_unknown	84.7%	100%of 14	Epoch AI Benchmarking Hub
ALE-Benchas gpt-5.5_xhigh	1943.0%	87%of 71	Epoch AI Benchmarking Hub
WeirdMLas gpt-5.5_xhigh	84.9%	86%of 52	Epoch AI Benchmarking Hub
SciCodeas gpt-5.5_high	56.1%	73%of 86	Epoch AI Benchmarking Hub
DeepSWEas gpt-5.5_xhigh	67.0%	60%of 26	Epoch AI Benchmarking Hub
FrontierCodeas gpt-5.5_unknown	43.0%	59%of 45	Epoch AI Benchmarking Hub
GSO (code optimisation)as gpt-5.5_xhigh	40.2%	30%of 11	Epoch AI Benchmarking Hub
Agents & real work · Agents 67
LMCAas gpt-5.5_xhigh	54.3%	82%of 67	Epoch AI Benchmarking Hub
DTBenchas gpt-5.5_xhigh	96.0%	81%of 67	Epoch AI Benchmarking Hub
Game Boy Advance evalas gpt-5.5_unknown	53.2%	77%of 23	Epoch AI Benchmarking Hub
DeepResearch Benchas gpt-5.5_high	54.0%	71%of 8	Epoch AI Benchmarking Hub
Vending-Bench 2as gpt-5.5_unknown	7523.8%	70%of 47	Epoch AI Benchmarking Hub
GDP.pdfas gpt-5.5_xhigh	26.0%	69%of 37	Epoch AI Benchmarking Hub
APEX-Agentsas gpt-5.5_unknown	55.1%	62%of 48	Epoch AI Benchmarking Hub
OSWorld 2.0as gpt-5.5_xhigh	13.0%	50%of 9	Epoch AI Benchmarking Hub
Remote Labor Indexas gpt-5.5_unknown	6.3%	38%of 9	Epoch AI Benchmarking Hub
Independent indexes · Indexes 83
Epoch Capabilities Indexas GPT-5.5	158.9	83%of 81	Epoch AI