GPT-6 Sol

+ contribute

GPT-6 Sol is a closed-weights language model from OpenAI, released September 2026.

82.6±0.7#7 of 719 LLMsfrom 20 benchmark results across 7 sources

$ how it compares

The LLMs ranked either side of GPT-6 Sol, out of 828 with a score, then more from OpenAI.

#	model	score	where	released
5	Claude Sonnet 5.5 Anthropic	84±1	api $4.00/M	2026-09
6	Claude Opus 5 Anthropic	83±1	api $10.00/M	2026-07
7	GPT-6 Sol OpenAI	83±1	api $4.00/M	2026-09
8	GPT-5.5 Pro OpenAI	82±1	api $67.50/M	2026-04
9	Gemini 4 Argon (limited preview) Google	82±1	limited preview	2026-09
	more from OpenAI
1	GPT-6 Astra OpenAI	87±1	api $20.00/M	2026-09
3	GPT-6.1 Sol OpenAI	85±1	api $4.00/M	2026-09
10	GPT-5.6 Sol OpenAI	82±1	api $8.00/M	2026-07

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▃ #7 of 300 · 21% of its score · 4 tests

test	result	vs the others in its run	difficulty	counts	why
Arena Agent (October 2026 aggregate IPS)	0.0971 IPS ✎	▁▁▁▄▂▃▇█▂▂▃▃ beats 44/49	avg -0.312 IPS	▇▇▇▇ 9.3%	sharp
Terminal-Bench 4.0	49.4% ✎	█▅▅▅▅·▅▃▃▅▅▅ beats 15/20	very hard · avg 0.4%	▇▇▇ 7.1%	sharp · at its level
DeepSWE v1.1 (resolved) lab	68.8% ✎	█··▃▃····▃·▃ beats 3/6	hard · avg 21%	▇ 2.4%	at its level
Agents' Last Exam (metric/harness unspecified) lab	56.4% ✎	▂·▂·█▂···▂▄▄ beats 10/11	very hard · avg 11%	▇ 1.8%	at its level

coding▁▁▁▃▆██▇▆▄▃▁ #7 of 458 · 13% of its score · 3 tests

test	result	vs the others in its run	difficulty	counts	why
Arena WebDev (Code Arena)	1689.0 ✎	▁▂▄▃▅█▄▇▆▄▁▁ beats 113/119	avg 981	▇▇▇ 8.2%	sharp
DeepSWE v1.1 (resolved) lab	68.8% ✎	█··▃▃····▃·▃ beats 3/6	hard · avg 21%	▇ 2.4%	at its level
Luau Leetcode fully solved (Torpedo Software)	82.7% ✎	▁▁▁▁·▁▂▃▅▃█▇ beats 39/49	hard · avg 18%	▇ 2.0%	typical

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #43 of 393 * · 4.6% of its score · 1 test

Trails MiMo-V2.5-Pro (1T-A42B), GLM-5.1, Grok 4.20 (0309 reasoning) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LMArena (text)	1457.2 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 328/379	avg 1268	▇▇ 4.6%	blunt

math▁▂▃▅▆▅▄█▄▁▃▁ #5 of 398 · 25% of its score · 3 tests

Beats Claude Sonnet 5.5, Claude Opus 5 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
FrontierMath tier 4 v2	90.0% ✎	▁▂▁▂█▅▂▂▃▁▁▁ beats 59/63	very hard · avg 0.1%	▇▇▇▇▇▇ 16%	sharp
FrontierMath tiers 1–3 v2	89.8% ✎	▂··▁▁▃▄▄█▅▃▃ beats 78/82	very hard · avg 1.2%	▇▇▇▇ 10%	sharp
OTIS mock AIME 2024–2025	100.0% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 185/193	very hard · avg 8.5%	▇ 0.2%	sharp · maxed out · noisy

other tests10 not in a task · 39% of its score

test	result	vs the others in its run	difficulty	counts	why
EBR-bench (Epoch)	53.3% ✎	▄▂·▆█▂▆▄▆▆·▄ beats 19/23	very hard · avg 0.3%	▇▇▇▇▇ 14%	sharp · at its level
Mystery Game Puzzles (Epoch)	56.0% ✎	█··▂▃▇▄█▂▃▁▁ beats 66/74	very hard · avg 9.4%	▇▇▇▇ 9.2%	sharp · at its level · noisy
Furniture Assembly (Epoch)	58.3% ✎	█▆▆▆▆▂▅▂▂▂·▅ beats 24/30	very hard · avg 1.2%	▇▇▇ 8.5%	sharp · at its level
SimpleQA Verified	60.7% ✎	▁▁▃▂▃▃▆▆█▃▂▅ beats 65/78	very hard · avg 8.0%	▇ 1.7%	at its level
Icelandic LLM Leaderboard average (Miðeind, Sept 2026 suite)	65.6% ✎	▁▁▃▂▂▄▁▃▄▄▄█ beats 32/38	very hard · avg 14%	▇ 1.5%	at its level
GPQA Diamond	94.3% ✎	▁▁▁▁▂▄▆▄▅▇█▅ beats 201/208	hard · avg 40%	▇ 1.5%	sharp · maxed out
HealthBench Hard lab	22.1% ✎	▃··▂····▂▄·█ beats 0/8	very hard · avg 9.7%	▇ 1.2%	blunt · at its level
HealthBench lab	47.1% ✎	▃·▃·····█▃▃▅ beats 0/8	hard · avg 24%	▇ 1.0%	blunt · at its level

2 more tests

test	result	vs the others in its run	difficulty	counts	why
RobloxQA v2.0 multiple choice (Torpedo Software)	97.8% ✎	▁▁▃▁█▄▃▃▆▂▃▂ beats 44/49	easy · avg 78%	▇ 0.7%	maxed out
Vectara hallucination leaderboard (HHEM-2.3, factual consistency rate)	93.5% ✎	▁▂▁▂▅█▂▃▄▃▁▁ beats 73/98	very easy · avg 90%	▇ 0.0%	blunt

$ where to run it

Through an API

api · $4.00/M · $2 in / $10 out per 1M tokens · 1.1M context · platform.openai.com ↗ · price sourceStandard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ.

$ how the score was worked out · Epoch AI — internal evaluatio… 60%, LMArena leaderboard dataset 22%, Terminal-Bench 4.0 official l… 7% and 4 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

source	results	share of the score	licence
Epoch AI — internal evaluations	8	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 60%	CC BY 4.0
LMArena leaderboard dataset	3	▇▇▇▇▇▇▇ 22%	CC BY 4.0
Terminal-Bench 4.0 official leaderboard submissions	1	▇▇ 7%	Apache-2.0
OpenAI vendor-reported benchmark scores for its own models (openai.com announcement posts and Deployment Safety Hub system cards) · lab-reported	4	▇▇ 6%	Factual vendor-reported score (not licensed content)
Roblox LLM Leaderboard (Torpedo Software): RobloxQA v2.0 and Luau Leetcode	2	▇ 3%	MIT
Miðeind Icelandic LLM Leaderboard	1	▇ 2%	CC BY 4.0
Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository)	1	▇ 0%	Apache-2.0
prior · pull toward the middle		▇ 0%
weighted average		= 82.6