Model comparison

Gemini 2.5 Pro vs GPT-5.6 Sol

GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 45.0 on the Noometry Index. Gemini 2.5 Pro costs 2.3× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.

Last verified October 11, 2026. 44 shared benchmarks.

Gemini 2.5 Pro

Google

45.0

Rank #75 Confirmed

GPT-5.6 Sol OpenAI

65.0

Rank #7 Confirmed

Summary

They share 44 benchmarks with published results for both. Gemini 2.5 Pro scores higher in 1 category and GPT-5.6 Sol in 9 categories; 9 gaps are clear of the uncertainty.
The widest gap is in math, where GPT-5.6 Sol leads 85.6 to 32.5.
The biggest single-benchmark swing is ARC-AGI-2: 4.9% for Gemini 2.5 Pro and 92.5% for GPT-5.6 Sol.
Gemini 2.5 Pro is cheaper at $1.25 / $10 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
GPT-5.6 Sol accepts more context: 1.05M tokens versus 1.05M.

Side by side

Gemini 2.5 Pro and GPT-5.6 Sol specifications
	Gemini 2.5 Pro	GPT-5.6 Sol
Provider	Google	OpenAI
Noometry Index	45.0	65.0
Released	2025-03-25	2026-07-09
Weights	Proprietary	Proprietary
Context window	1.05M	1.05M
Max output	66K	128K
Input $ / M tokens	$1.25	$4
Output $ / M tokens	$10	$20
Results tracked	78	65

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding GPT-5.6 Sol leads

Gemini 2.5 Pro: 42.4 (#101), GPT-5.6 Sol: 65.1 (#7)

Coding benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
LMArena WebDev	1227	1618
SciCode	42.8%	57.1%
GSO	3.9%	76.5%
WeirdML	54%	89.4%
LMArena Coding	1452	1498
ALE-Bench	785.52	2,177
SWE-bench Verified	57.6%	—
DeepSWE	—	72.7%
FrontierCode	—	47.5%
SWE-bench Verified (bash only)	53.6%	—
Aider Polyglot	83.1%	—
CursorBench	—	41.7%
FrontierSWE	—	32.2%
LiveBench Coding	85.9%	—
MirrorCode	—	20%
CadEval	64%	—
AlgoTune	1.51	—

Agentic & Tool Use GPT-5.6 Sol leads

Gemini 2.5 Pro: 29.2 (#88), GPT-5.6 Sol: 50.3 (#7)

Agentic & Tool Use benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
τ²-bench Banking	13.7%	46.9%
BALROG	43.3%	60%
LMArena Search	1142	1257
Vending-Bench 2	573.64	9,619
Terminal-Bench	32.6%	—
APEX-Agents	—	51.4%
OSWorld 2.0	—	27.3%
GDPval	23.3%	—
Remote Labor Index	0.8%	—
TheAgentCompany	30.3%	—
DeepResearch Bench	42.8%	—
PostTrainBench	—	36.2%
GBAEval	—	52.6%
GDP.pdf	—	30.7%
METR Time Horizons	55.4%	—

Reasoning GPT-5.6 Sol leads

Gemini 2.5 Pro: 28.8 (#99), GPT-5.6 Sol: 74.8 (#8)

Reasoning benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
ARC-AGI-2	4.9%	92.5%
SimpleBench	62.4%	71.7%
Kagi LLM Benchmark	70.3%	67%
ARC-AGI-1	41%	97.5%
CritPt	2%	32.3%
Chess Puzzles	20%	64%
EnigmaEval	5.6%	37.1%
LMArena Hard Prompts	1455	1484
DTBench	82.4%	96%
LMCA	34.8%	59.2%
Epoch Capabilities Index	145.32	161.66
NYT Connections (extended)	—	93.8%
EBR-Bench	—	44.8%
LiveBench Reasoning	89.8%	—
Mystery Game Puzzles	—	58%
LiveBench Data Analysis	79.9%	—
Surface Evolver Bench	—	93.1%
Bench to the Future 3	—	0.14
ForecastBench	61.3	—
LiveBench	82.3%	—

Math GPT-5.6 Sol leads

Gemini 2.5 Pro: 32.5 (#213), GPT-5.6 Sol: 85.6 (#9)

Math benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
FrontierMath (Tiers 1-3)	24.6%	89.1%
FrontierMath Tier 4	0%	82.9%
OTIS Mock AIME 2024-2025	84.7%	100%
LMArena Math	1450	1474
ProofBench	—	83%
Omni-MATH	41.6%	—
LiveBench Math	90.2%	—
MATH Level 5	95.9%	—
FrontierMath (Feb 2025 set)	14.1%	—
FrontierMath Erdős	—	0%
FrontierMath Tier 4 (v1)	4.2%	—

Knowledge GPT-5.6 Sol leads

Gemini 2.5 Pro: 56.0 (#46), GPT-5.6 Sol: 64.3 (#18)

Knowledge benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
GPQA Diamond	85.3%	93.5%
Vectara Hallucination Rate	7%	12.4%
LMArena Expert	1452	1516
Humanity's Last Exam	21.6%	—
SimpleQA Verified	—	69.7%
MMLU-Pro	86.3%	—
Confabulations	10.6%	—
GPQA (HELM)	74.9%	—

Multimodal GPT-5.6 Sol leads

Gemini 2.5 Pro: 45.2 (#18), GPT-5.6 Sol: 48.6 (#9)

Multimodal benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
LMArena Vision	1263	1281
LMArena Document	1421	1483
GeoBench	86%	—
VPCT	48%	—
Blueprint-Bench 2	—	33.6%
Furniture Assembly	—	56.7%
SpatialViz-Bench	44.7%	—

Multilingual Too close to call

Gemini 2.5 Pro: 55.3 (#31), GPT-5.6 Sol: 55.3 (#32)

Multilingual benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
LMArena Non-English	1451	1452
LMArena Chinese	1507	1527
LMArena French	1472	1477
LMArena German	1487	1476
LMArena Japanese	1461	1471
LMArena Korean	1434	1442
LMArena Russian	1461	1468
LMArena Spanish	1473	1441

Instruction Following GPT-5.6 Sol leads

Gemini 2.5 Pro: 75.0 (#75), GPT-5.6 Sol: 77.7 (#16)

Instruction Following benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
LMArena Instruction Following	1437	1482
LiveBench Instruction Following	80.6%	—
IFEval	84%	—

Long Context Gemini 2.5 Pro leads

Gemini 2.5 Pro: 59.8 (#5), GPT-5.6 Sol: 45.4 (#42)

Long Context benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
LMArena Longer Query	1449	1480
Fiction.LiveBench	91.7%	—

Writing & Preference GPT-5.6 Sol leads

Gemini 2.5 Pro: 63.7 (#62), GPT-5.6 Sol: 73.3 (#12)

Writing & Preference benchmarks
Benchmark	Gemini 2.5 Pro	GPT-5.6 Sol
LMArena Text	1458	1457
LMArena Creative Writing	1454	1448
EQ-Bench Creative Writing	1421	1972
LMArena Multi-Turn	1453	1460
Short-Story Creative Writing	83.8%	—
WildBench	85.7%	—
EQ-Bench 4	—	1250
LiveBench Language	67.8%	—

Frequently asked questions

Is Gemini 2.5 Pro better than GPT-5.6 Sol?

Which is cheaper, Gemini 2.5 Pro or GPT-5.6 Sol?

Gemini 2.5 Pro is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; GPT-5.6 Sol lists at $4 and $20.

Is Gemini 2.5 Pro or GPT-5.6 Sol better for coding?

GPT-5.6 Sol scores higher on coding benchmarks: 65.1 versus 42.4 in the Noometry coding category.

Which has the bigger context window?

GPT-5.6 Sol does, with 1.05M tokens against 1.05M.

How many benchmarks do Gemini 2.5 Pro and GPT-5.6 Sol share?

44 benchmarks have published results for both models. Gemini 2.5 Pro has 78 scored results on Noometry and GPT-5.6 Sol has 65.