whatmodel.sh / Claude Sonnet 5.5

Claude Sonnet 5.5

+ contribute

Claude Sonnet 5.5 is a closed-weights language model from Anthropic, released September 2026. Licence: proprietary.

Catalogue identity notes

Official claude-sonnet-5-5, released September 28, 2026. Arena xhigh setting; lower-effort entries are not averaged into its score.

Model information sources

Release date ↗

83.9±0.7#5 of 719 LLMsfrom 26 benchmark results across 6 sources

$ how it compares

The LLMs ranked either side of Claude Sonnet 5.5, out of 828 with a score, then more from Anthropic.

#	model	score	where	released
3	GPT-6.1 Sol OpenAI	85±1	api $4.00/M	2026-09
4	Claude Fable 5.1 Anthropic	85±1	api $20.00/M	2026-09
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
	more from Anthropic
2	Claude Opus 5.5 Anthropic	87±1	api $8.00/M	2026-09
11	Claude Fable 5 Anthropic	81±1	api $20.00/M	2026-06
12	Claude Mythos Preview (limited preview) Anthropic	81±1	limited preview	2026-04

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▃ #2 of 300 · 39% of its score · 8 tests

Beats GPT-6 Astra, GPT-6.1 Sol, Claude Fable 5.1 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Terminal-Bench 4.0 +1	61.8% ✎	█▅▅▅▅·▅▃▃▅▅▅ beats 19/20	very hard · avg 0.5%	▇▇▇▇▇ 12%	sharp · at its level · several runs
Arena Agent (October 2026 aggregate IPS)	0.1252 IPS ✎	▁▁▁▄▂▃▇█▂▂▁▃ beats 47/49	avg -0.312 IPS	▇▇▇ 8.0%	sharp
SWE-bench Pro (split unspecified) lab	81.3% ✎	▄█·▄·▄██···▄ beats 8/9	very hard · avg 7.6%	▇▇▇ 6.7%	sharp
SWE-bench Multilingual lab	90.3% ✎	▃··▃·▅▃▃█··▃ beats 8/9	hard · avg 31%	▇ 3.2%	typical
Chartography with tools lab	90.2% ✎	▃········▃██ beats 7/7	very hard · avg 4.7%	▇ 2.6%	sharp
Humanity's Last Exam with tools lab	64.5% ✎	▁······▁·▄·█ beats 7/10	very hard · avg 14%	▇ 2.3%	at its level
DeepSWE v1.1 (resolved) lab	71.0% ✎	█·····████·█ beats 4/5	hard · avg 18%	▇ 2.1%	at its level
Toolathlon-Verified lab	77.8% ✎	▃·····█···▃▃ beats 1/5	hard · avg 35%	▇ 1.1%	blunt

coding▁▁▁▃▆██▇▆▄▃▃ #3 of 458 · 21% of its score · 5 tests

Beats GPT-6.1 Sol, Claude Fable 5.1 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Arena WebDev (Code Arena)	1786.3 ✎	▁▂▄▃▅█▄▇▆▄▁▃ beats 117/119	avg 981	▇▇▇ 7.1%	sharp
SWE-bench Pro (split unspecified) lab	81.3% ✎	▄█·▄·▄██···▄ beats 8/9	very hard · avg 7.6%	▇▇▇ 6.7%	sharp
SWE-bench Multilingual lab	90.3% ✎	▃··▃·▅▃▃█··▃ beats 8/9	hard · avg 31%	▇ 3.2%	typical
DeepSWE v1.1 (resolved) lab	71.0% ✎	█·····████·█ beats 4/5	hard · avg 18%	▇ 2.1%	at its level
Luau Leetcode fully solved (Torpedo Software)	80.8% ✎	▁▁▁▁·▁▂▃▅▃█▇ beats 35/49	hard · avg 18%	▇ 1.6%	typical

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #34 of 393 * · 4.0% of its score · 1 test

Trails Grok 4.20 (0309 reasoning), Claude Opus 4.5, Gemini 3 Flash (Preview) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LMArena (text)	1471.0 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 342/379	avg 1268	▇▇ 4.0%	blunt

math▁▂▃▅▆▅▄█▄▁▃▁ #8 of 398 · 19% of its score · 4 tests

Trails Claude Fable 5, GPT-5.6 Sol, GPT-6 Sol here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
FrontierMath tier 4 v2	80.5% ✎	▁▂▁▂█▅▂▃▁▁▁▁ beats 56/63	very hard · avg 0.1%	▇▇▇▇ 11%	sharp
FrontierMath tiers 1–3 v2	88.8% ✎	▂··▁▁▃▄▄█▅▃▃ beats 76/82	very hard · avg 1.2%	▇▇▇ 6.9%	sharp
FrontierMath: Erdős problems	2.9% ✎	█··········▆ beats 4/6	very hard · avg 0.0%	▇ 0.7%	sharp · too hard for it · noisy
OTIS mock AIME 2024–2025	100.0% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 186/193	very hard · avg 8.5%	▇ 0.1%	sharp · maxed out · noisy

vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #32 of 147 * · 2.8% of its score · 1 test

Trails Qwen3.6 Plus, Qwen3.7 Plus, Gemini 3 Flash (Preview) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena vision (style controlled)	1268.3 ✎	▁▁▂▂▂▂▄▅▅▆██ beats 117/145	avg 1068	▇ 2.8%	blunt

web search▂▁▂▆▄█▃▃▁▁·▁ #18 of 80 * · 2.3% of its score · 1 test

Trails GPT-4o mini (2024-07-18), Mistral Small 3.1 24B (2503), GPT-4o (2024-08-06) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Humanity's Last Exam with tools lab	64.5% ✎	▁······▁·▄·█ beats 7/10	very hard · avg 14%	▇ 2.3%	at its level

other tests9 not in a task · 27% of its score

test	result	vs the others in its run	difficulty	counts	why
Mystery Game Puzzles (Epoch)	65.0% ✎	█··▂▃▇▄█▂▃▃▁ beats 71/74	very hard · avg 9.4%	▇▇▇ 7.8%	sharp · at its level · noisy
Furniture Assembly (Epoch)	75.0% ✎	█▆▆▆▆▂▅▂▂▃·▅ beats 27/30	very hard · avg 1.2%	▇▇▇ 7.1%	sharp · at its level
Humanity's Last Exam lab	56.9% ✎	▂·····▂·▄▂█▄ beats 7/10	very hard · avg 5.8%	▇▇ 4.9%	at its level
Global-MMLU lab	92.1% ✎	▄▄····▄▄··▄█ beats 2/6	medium · avg 63%	▇ 1.7%	typical
SimpleQA Verified	46.5% ✎	▁▁▃▂▃▃▆▆█▃▂▅ beats 45/78	very hard · avg 8.0%	▇ 1.4%	at its level
Icelandic LLM Leaderboard average (Miðeind, Sept 2026 suite)	62.1% ✎	▁▁▃▂▂▄▁▃▄▄▄█ beats 26/38	very hard · avg 14%	▇ 1.3%	at its level
GPQA Diamond	95.6% ✎	▁▁▁▁▂▄▆▄▅▇█▅ beats 207/208	hard · avg 40%	▇ 1.1%	sharp · maxed out
HealthBench lab	69.4% ✎	█·▄·····▄▄▄▄ beats 6/6	hard · avg 29%	▇ 1.0%	blunt · at its level

1 more test

test	result	vs the others in its run	difficulty	counts	why
RobloxQA v2.0 multiple choice (Torpedo Software)	96.6% ✎	▁▁▃▁█▄▃▃▆▂▂▂ beats 37/49	easy · avg 78%	▇ 0.6%	maxed out

$ where to run it

Through an API

api · $4.00/M · $2 in / $10 out per 1M tokens · 1M context · platform.claude.com ↗ · price sourceObserved 2026-10-06. USD per million tokens, standard uncached input/output; batch and cache discounts excluded.

$ how the score was worked out · Epoch AI — internal evaluatio… 36%, Anthropic system cards and mo… 32%, LMArena leaderboard dataset 22% and 3 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	▇▇▇▇▇▇▇▇▇▇▇ 36%	CC BY 4.0
Anthropic system cards and model announcements (lab-reported scores) · lab-reported	10	▇▇▇▇▇▇▇▇▇▇ 32%	Factual vendor-reported score (not licensed content)
LMArena leaderboard dataset	4	▇▇▇▇▇▇▇ 22%	CC BY 4.0
Terminal-Bench 4.0 official leaderboard submissions	1	▇▇ 6%	Apache-2.0
Roblox LLM Leaderboard (Torpedo Software): RobloxQA v2.0 and Luau Leetcode	2	▇ 2%	MIT
Miðeind Icelandic LLM Leaderboard	1	▇ 1%	CC BY 4.0
prior · pull toward the middle		▇ 0%
weighted average		= 83.9