Claude Sonnet 4.5

Claude Sonnet 4.5 ranks first on 4 of the 204 benchmarks it has scores on.

Anthropic Released 29 Sep 2025 Family Claude 4.5 The lab's model card

208

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Anthropic Sep 2025

Claude Sonnet 4.5

Closed weights Mid-size

146.8 #78 of 271

Epoch Capabilities Index

Humanity's Last Exam 13.7% #25 /47
SWE-bench Verified 71.3% #23 /32
SWE-Bench Pro 43.6% #7 /24

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

Benchmark	Score	Rank	Source
MMLU-Pro	87.4%	#10 of 212	Official leaderboard
Phare	77.6%	#4 of 62	Official leaderboard
WritingBench	80.7	#4 of 45	Official leaderboard
BullshitBench V2	79%	#12 of 122	Official leaderboard
BullshitBench	65.5%	#11 of 107	Official leaderboard
Slop Score	19.5	#2 of 17	Official leaderboard
IndicGenBench	60.6	#2 of 13	Official leaderboard
LMArena (Elo)	1,456	#52 of 333	Official leaderboard
Judgemark v2.1	84.6	#8 of 42	Official leaderboard
EQ-Bench 3	1,535	#15 of 78	Official leaderboard
MMMLU	89.1%	#13 of 62	Model card
Longform Creative Writing (EQ-Bench)	71.4	#38 of 119	Official leaderboard
False-premise questions (Anthropic)	6.90%	#2 of 6	Model card
Vectara Hallucination Leaderboard (HHEM)	5.50%	#50 of 148	Official leaderboard
Creative Writing v3	72.2	#43 of 120	Official leaderboard
MultiNRC	35.8%	#17 of 38	Official leaderboard
FACTS Search	69.8%	#14 of 31	Official leaderboard
SimpleQA	24.6%	#31 of 59	Model card
Humanity's Last Exam	13.7%	#25 of 47	Official leaderboard
SimpleQA Verified	29.2%	#32 of 56	Official leaderboard
100Q-Hard (Anthropic)	38.3%	#4 of 7	Model card
FACTS Parametric	29%	#29 of 45	Official leaderboard
Honesty human feedback win rate (Anthropic)	57%	#4 of 6	Model card
MultiChallenge	55.3%	#25 of 35	Official leaderboard
Extended NYT Connections	37.3%	#72 of 97	Official leaderboard
Document Arena	1,449	#28 of 37	Official leaderboard
Vectara Hallucination Leaderboard (2025 dataset)	12%	#79 of 104	Official leaderboard
FACTS Multimodal	31.1%	#30 of 39	Official leaderboard
FACTS Benchmark Suite	31.1%	#36 of 40	Official leaderboard

Math and science

Benchmark	Score	Rank	Source
AIME 2025	100%	#4 of 143	Model card
MATH Level 5	97.7%	#5 of 94	Epoch AI
CORE-Bench	62.2%	#2 of 17	Official leaderboard
GPQA Diamond	82.3%	#82 of 219	Epoch AI
DTBench	72%	#74 of 178	Official leaderboard
OTIS Mock AIME 2024-2025	77.8%	#91 of 202	Epoch AI
SimpleBench	54.3%	#47 of 100	Official leaderboard
FrontierMath	15.2%	#34 of 70	Epoch AI
BRUMO 2025	90.8%	#21 of 42	Official leaderboard
ProofBench	19%	#31 of 59	Official leaderboard
ARC-AGI-2	13.6%	#54 of 95	Official leaderboard
CritPt	1.14%	#91 of 160	Epoch AI
FrontierMath Tier 4	4.17%	#32 of 55	Epoch AI
ARC-AGI	63.7%	#55 of 92	Official leaderboard
HMMT 2025	67.5%	#33 of 53	Official leaderboard
SMT 2025	84%	#25 of 40	Official leaderboard
AMO-Bench	17.6%	#14 of 21	Official leaderboard
ForecastBench	64.0	#59 of 86	Official leaderboard
MathArena (Apex, IMO, USAMO)	1.56%	#34 of 47	Official leaderboard
General365	48.6%	#18 of 23	Official leaderboard
FrontierMath Tiers 1-3 v2	23.9%	#74 of 89	Epoch AI
CMIMC 2025	66.3%	#30 of 33	Official leaderboard
FrontierMath Tier 4 v2	2.44%	#66 of 71	Epoch AI

Coding

Benchmark	Score	Rank	Source
LiveSQLBench	35.2%	#4 of 36	Official leaderboard
CodeClash	1,385	#1 of 8	Official leaderboard
METR Time Horizon	117 min	#4 of 31	Official leaderboard
OTelBench	21.7%	#3 of 14	Official leaderboard
SWE-Bench Pro	43.6%	#7 of 24	Official leaderboard
SWE-fficiency	11.6%	#5 of 17	Official leaderboard
Vibe Code Bench v1	22.6%	#5 of 17	Official leaderboard
Bird-SQL	66.8%	#4 of 13	Official leaderboard
Time series forecasting task (Anthropic)	5.30	#3 of 9	Model card
IOI (Vals)	18.3%	#24 of 55	Official leaderboard
SWE-bench Verified (hard subset)	45.3%	#4 of 9	Model card
OpenHands Index	53%	#15 of 33	Official leaderboard
SWE-bench-Live	36%	#6 of 13	Official leaderboard
WeirdML	47.7%	#52 of 112	Official leaderboard
GSO-Bench	14.7%	#15 of 30	Official leaderboard
CooperBench	25.9%	#3 of 6	Official leaderboard
AlgoTune	1.52×	#10 of 18	Official leaderboard
Text-based reinforcement learning task (Anthropic)	0.85	#4 of 7	Model card
ALE-Bench	796	#71 of 121	Epoch AI
SWE-bench Multilingual	67%	#8 of 13	Official leaderboard
Quadruped reinforcement learning task (Anthropic)	1.30	#5 of 8	Model card
WebDev Arena	1,391	#70 of 105	Epoch AI
Internal AI research evaluation suite 2	0.51	#4 of 6	Model card
Vibe Code Bench	22.6%	#79 of 115	Official leaderboard
Code Arena	1,394	#86 of 124	Official leaderboard
SWE-bench Verified	71.3%	#23 of 32	Epoch AI
PostTrainBench	9.94%	#18 of 23	Official leaderboard
Internal model use survey (Anthropic)	0%	#4 of 5	Model card

Agents and tools

Benchmark	Score	Rank	Source
BFCL V4	73.2%	#2 of 65	Official leaderboard
AgentBench FC	58.9%	#1 of 12	Official leaderboard
SnakeBench	33	#19 of 183	Official leaderboard
Terminal-Bench	46.5%	#2 of 18	Epoch AI
MCP-Universe	35.1%	#4 of 30	Official leaderboard
Elimination Game	5	#11 of 56	Official leaderboard
DeepResearch Bench	52.6%	#5 of 24	Epoch AI
MCPMark	32.1%	#6 of 28	Official leaderboard
Step Race	3	#14 of 59	Official leaderboard
GAIA	73.4%	#15 of 57	Official leaderboard
LisanBench	2665.3	#38 of 133	Official leaderboard
CompileBench	86.7%	#6 of 20	Official leaderboard
VitaBench 2.0	41.7%	#6 of 20	Official leaderboard
BalatroBench	11.9	#5 of 16	Official leaderboard
tau2-bench	84.7%	#26 of 69	Model card
PinchBench	71.3%	#19 of 48	Official leaderboard
Toolathlon	38.9%	#21 of 50	Official leaderboard
OSWorld-Verified	62.9%	#14 of 33	Official leaderboard
Terminal-Bench 2.0	46.5%	#19 of 43	Official leaderboard
VITA-Bench	23.5%	#8 of 18	Official leaderboard
OSWorld	62.9%	#4 of 9	Epoch AI
Agent Poker Bench	1,056	#8 of 17	Official leaderboard
Pencil Puzzle Bench	3.30%	#25 of 53	Official leaderboard
LLM Chess	113	#60 of 124	Official leaderboard
Game Arena Coin Game	355	#18 of 35	Official leaderboard
Game Arena Poker	-0.72	#17 of 31	Official leaderboard
Game Arena Clobber	225	#21 of 36	Official leaderboard
ITBench	45.8%	#8 of 13	Official leaderboard
Game Arena Bargaining	167	#24 of 36	Official leaderboard
MobileWorld	47.8%	#12 of 18	Official leaderboard
Mystery Game Puzzles	17%	#54 of 80	Epoch AI
τ³-bench Banking	25.3%	#17 of 25	Official leaderboard
Game Arena Word Association	128	#25 of 36	Official leaderboard
Game Arena Ultimate Tic-Tac-Toe	178	#26 of 36	Official leaderboard
Game Arena Lines of Action	344	#28 of 36	Official leaderboard
Game Arena Checkers	565	#30 of 38	Official leaderboard
Game Arena Word Art	218	#28 of 35	Official leaderboard
Game Arena Chess Openings	228	#9 of 11	Official leaderboard
Kaggle Game Arena	173	#32 of 39	Official leaderboard
MCP Atlas	59.5%	#28 of 34	Official leaderboard
LMArena Search Arena	1,176	#29 of 35	Official leaderboard
Game Arena Nine Men's Morris	356	#30 of 36	Official leaderboard
Game Arena Reversi	153	#30 of 36	Official leaderboard
SkillsBench	36.2%	#18 of 21	Official leaderboard
Game Arena Four in a Row	157	#33 of 38	Official leaderboard
Game Arena Go	325	#32 of 36	Official leaderboard
Game Arena Dots and Boxes	529	#33 of 37	Official leaderboard
DeepSearchQA	16%	#9 of 10	Official leaderboard
Chess Puzzles	10%	#75 of 83	Epoch AI
EBR-bench	2.38%	#27 of 28	Epoch AI
Game Arena Dark Hex	42	#37 of 38	Official leaderboard
Game Arena Chess	0	#39 of 39	Official leaderboard
Game Arena Five in a Row	0	#20 of 20	Official leaderboard
tau3-bench	76.4%	#7 of 7	Official leaderboard

Professional work

Benchmark	Score	Rank	Source
Vals Finance Agent v1	55.3%	#3 of 47	Official leaderboard
LEXam	62.8%	#4 of 31	Official leaderboard
LegalBench	84.1%	#39 of 138	Official leaderboard
Vals Finance Agent	54.5%	#16 of 49	Official leaderboard
MedScribe	84.5%	#33 of 98	Official leaderboard
CaseLaw v2	62.2%	#19 of 53	Official leaderboard
GDPval	42.5%	#4 of 11	Epoch AI
TaxEval v2	73.3%	#49 of 129	Official leaderboard
MedCode	44.1%	#40 of 96	Official leaderboard
CorpFin v2	62%	#50 of 119	Official leaderboard
Vending-Bench	$2,465	#4 of 9	Official leaderboard
ProfBench	48.1%	#36 of 72	Official leaderboard
MortgageTax	64%	#46 of 86	Official leaderboard
FailureSensorIQ	64.2%	#13 of 24	Official leaderboard
Vending-Bench 2	$3,839	#42 of 66	Official leaderboard
PRBench	40.8%	#31 of 44	Official leaderboard
TaxCalcBench	31.4%	#12 of 17	Official leaderboard
Vending-Bench Arena	$2,602	#23 of 32	Official leaderboard
Remote Labor Index	2.08%	#13 of 17	Official leaderboard
prinzbench	20.0	#23 of 26	Official leaderboard
JobBench	20.7%	#15 of 16	Official leaderboard

Vision, audio and robotics

Benchmark	Score	Rank	Source
MMMU	77.8%	#14 of 136	Official leaderboard
MADQA	80.6%	#2 of 16	Official leaderboard
MATH-Vision (MathVision)	71.1%	#25 of 105	Official leaderboard
VISTA	48.8%	#14 of 54	Official leaderboard
MMMU-Pro	68.9%	#15 of 57	Official leaderboard
EnigmaEval	6.00%	#20 of 41	Epoch AI
OCR Arena	1,631	#6 of 11	Official leaderboard
VPCT	39.8%	#17 of 28	Epoch AI
ZeroBench	5.00%	#43 of 66	Official leaderboard
Math Kangaroo 2025 (MathArena)	75.8%	#16 of 22	Official leaderboard
GeoBench	60%	#36 of 48	Official leaderboard
ClockBench	7.80%	#33 of 37	Official leaderboard
VoxelBench	1,246	#53 of 56	Official leaderboard

Safety and security

Benchmark	Score	Rank	Source
AIR-Bench 2024	89.8%	#3 of 82	Official leaderboard
Spiral-Bench v1.2	70.3	#2 of 21	Official leaderboard
FORTRESS	12.8%	#7 of 66	Official leaderboard
Agentic misalignment suite (Anthropic)	0%	#1 of 7	Model card
HELM Safety	97.1	#11 of 75	Official leaderboard
Single-turn benign request evaluations (Anthropic)	0.02%	#3 of 18	Model card
KORA	63.6%	#8 of 42	Official leaderboard
Computer use prompt injection evaluation (Anthropic)	78%	#1 of 5	Model card
Cybench	60%	#6 of 24	Official leaderboard
LAB-Bench CloningScenarios	66.7%	#3 of 11	Model card
Single-turn harmful request evaluations (Anthropic)	99.3%	#5 of 18	Model card
SimpleSafetyTests	100%	#22 of 75	Official leaderboard
Spiral-Bench v1.0	76.4	#4 of 13	Official leaderboard
BinaryAudit	26.3%	#8 of 20	Official leaderboard
LAB-Bench ProtocolQA	83.3%	#7 of 15	Model card
Reward-hack-prone coding tasks (Anthropic)	1.00%	#4 of 8	Model card
Tool use prompt injection evaluation (Anthropic)	96%	#1 of 2	Model card
Creative biology	48.8%	#4 of 7	Model card
LAB-Bench SeqQA	78%	#7 of 12	Model card
Average in-eval reward hacking rate (Anthropic)	15.4%	#3 of 5	Model card
BBQ (Bias Benchmark for QA)	82.2%	#15 of 24	Model card
CyberGym	28.9%	#22 of 33	Official leaderboard
LAB-Bench FigQA	49.7%	#7 of 10	Model card
Anthropic Red Team	98.2%	#54 of 75	Official leaderboard
SHADE-Arena	1.50%	#13 of 16	Model card
Virology Capabilities Test (VCT)	39.7%	#24 of 28	Model card
Gray Swan Agent Red Teaming (ART) benchmark	40%	#12 of 14	Model card
Impossible tasks (Anthropic)	53%	#10 of 11	Model card
Agentic coding malicious requests (Anthropic)	98.7%	#5 of 5	Model card
Training distribution hack rate (Anthropic)	2.00%	#3 of 3	Model card
Apollo Research strategic deception evaluations	0%	#2 of 2	Model card
MCP prompt injection evaluation (Anthropic)	92%	#2 of 2	Model card

Other

Benchmark	Score	Rank	Source
HELM Capabilities	76.2	#12 of 60	Official leaderboard
CAIS Risk Index	40.5	#15 of 65	Official leaderboard
ACCoRD	70.7	#48 of 150	Official leaderboard
LMCA	38.8%	#54 of 139	Epoch AI
Conceptual Reasoning Index	51.8	#62 of 149	Official leaderboard
CAIS Text Capabilities Index	22.9	#46 of 66	Official leaderboard
CAIS Vision Capabilities Index	42.1	#41 of 53	Official leaderboard
SAGE (Vals)	36.1%	#64 of 81	Official leaderboard

Anthropic's models over time

Anthropic's models in color, against every other model on the Epoch Capabilities Index, by release date.

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

Introducing Claude Sonnet 4.5 (launch post) 29 Sep 2025

Its system card

Every result in the longer documents the lab published with it. 56 of them aren't in the launch post.

Claude Sonnet 4.5

Claude Sonnet 4.5

Where it ranks

Knowledge and language

Math and science

Coding

Agents and tools

Professional work

Vision, audio and robotics

Safety and security

Other

Anthropic's models over time

Its launch post

Its system card

Hallucination

Safety

Robustness

Bias

Alignment

Bio and chem risk

Cyber risk

AI R&D risk