Claude 3.5 Sonnet

Claude 3.5 Sonnet is a non-reasoning model from Anthropic. 12 benchmarks count toward its score, in 4 categories.

availableShows if the model has enough results for an index.
IndexOverall score out of 100.31.8 ±9.2
CoverageShare of the index weight with results.55%
SpeedOutput tokens per second.
Input / 1MUS dollars per 1M input tokens.$3
Output / 1MUS dollars per 1M output tokens.$15
ContextMaximum tokens in one request.200K
EloLMArena rating and rank.1281 (#257)

The index is a score out of 100. The ± range shows how much it can change.

82,419 votes. Elo shows what people prefer. It does not change the score.

CapabilitiesScore per category, out of 100.

Out of 100
AgenticMulti-step tasks with tools.
N/A
CodingCode writing and repair.
33.2
ReasoningLogic problems and puzzles.
N/A
MultimodalTasks with images and text.
39.3
KnowledgeFacts and expert knowledge.
38.5
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
N/A
MathMath problems.
25.9

Results

12 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result as a score out of 100.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
MGSMMultilingual92.6%9 Jan 2026Vals AI
MMLU ProKnowledge78.4%44.01 Sept 2026Vals AI
MATH 500Math72.4%22.79 Jan 2026Vals AI
MMMU ProMultimodal68.8%39.31 Sept 2026Vals AI
Graduate-Level Google-Proof Q&AKnowledge59.4%33.1David Rein et al.
GPQA DiamondKnowledge59.3%33.01 Sept 2026Vals AI
GPQA diamondKnowledge54.0%28.1Epoch AI
SWE-bench VerifiedCoding53.0%40.0Engine Labs1 Sept 2026SWE-bench team
MATH level 5Math51.7%24.1Epoch AI
LiveCodeBenchCoding49.6%29.51 Sept 2026Vals AI
Software Engineering Benchmark VerifiedCoding49.0%36.8Carlos E. Jimenez et al.
SWE-bench LiteCoding40.8%38.1Agentless-1.511 Sept 2025SWE-bench team
SWE-bench VerifiedCoding33.6%24.5SWE-agent26 Feb 2026SWE-bench team
SWE-bench LiteCoding30.0%28.8Kortix AI11 Sept 2025SWE-bench team
SWE-bench (full test split)Coding21.5%SWE-agent19 Dec 2025SWE-bench team
AIMEMath10.0%18.416 Apr 2026Vals AI
SWE-bench MultimodalMultimodal8.8%Agentless17 Nov 2025SWE-bench team
SWE-bench MultimodalMultimodal8.8%SWE-agent Multimodal17 Nov 2025SWE-bench team
OTIS Mock AIME 2024-2025Math6.5%15.3Epoch AI
FrontierMath-2025-02-28-PrivateMath1.0%31.6Epoch AI
FrontierMath-Tier-4-2025-07-01-PrivateMath0.0%43.4Epoch AI

12 benchmarks count, from 17 of 21 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authorsVals AI, collected directlyNo licence stated. Read from the public leaderboard and credited to Vals AIEpoch AI, collected directlyCC BY — free to use and redistribute with attributionSWE-bench team, collected directlyNo licence stated. The repository publishes submission records for reproducibility and transparency and asks that SWE-bench be cited

Same level, lower price

Qwen3.8-Flash-Next66.4 · Freedots3-note Preview65.5 · FreeApodex 1.1 Mini59.1 · Free

More from Anthropic

Claude Opus 5.582.8Claude Fable 5.181.4Claude Opus 577.9Claude Fable 577.5Claude Opus 4.871.1