Anthropic · model passport
Claude 3 Sonnet
claude-3-sonnet- Cloud only
- Proprietary licence
- Released 2024-03-04(curated)
- Overall · LMArena
- 1,281#282 of 396
- Best percentile
- 30.4Coding
- Cohort standing
- 29.4median 31.7
What the data says
Benchmarked on 10 of 27 tracked axes. Every figure below is the model's own published result on that board.
Weakest axes
- 9.3 pct — Vision Arena (1,016)
- 11 pct — AIME (math) (2.5%)
Uneven: 21.1 percentile points between its best and worst axis — 30.4 pct on Coding versus 9.3 pct on Vision Arena.
- Coding30.4 pct1,318Verify →
- Instruction following30.1 pct1,268Verify →
- Overall29 pct1,281Verify →
- Hard prompts29 pct1,283Verify →
- MMLU-Pro (Artificial Analysis)25.1 pct57.9%Verify →
- LiveBench23.7 pct38.1%Verify →
- LiveCodeBench (Artificial Analysis)22.5 pct17.5%Verify →
- GPQA Diamond21.7 pct40.6%Verify →
- AIME (math)11 pct2.5%Verify →
- Vision Arena9.3 pct1,016Verify →
- WebDev ArenaNot benchmarked
- Factuality ArenaNot benchmarked
- Search ArenaNot benchmarked
- Document ArenaNot benchmarked
- Agent ArenaNot benchmarked
- FrontierMathNot benchmarked
- SimpleQA VerifiedNot benchmarked
- ARC-AGI-2Not benchmarked
- Humanity's Last ExamNot benchmarked
- SimpleBenchNot benchmarked
- BFCL v4 (function calling)Not benchmarked
- SWE-bench Verified (live board)Not benchmarked
- Terminal-Bench 4.0Not benchmarked
- IFBench (Artificial Analysis)Not benchmarked
- HHEM Hallucination (Vectara)Not benchmarked
- SWE-bench VerifiedNot benchmarked
- Terminal-BenchNot benchmarked
Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.
claude-3-sonnet-20240229 stands at 29.4 (raw mean 23.2) — measured on 10 of 27 axes.
claude-3-sonnet-20240229 stands at 29.4 versus a cohort median of 31.7 across 102 models released within 180 days — below the peer median.
On OTIS Mock AIME, claude-3-sonnet-20240229 raised the running best score by 0.3 percentage points to 2.5% on 2024-03-04.
claude-3-sonnet-20240229 scores at percentile 9.3 on Vision Arena.
claude-3-sonnet-20240229 scores in the 11th percentile on AIME (math).
claude-3-sonnet-20240229 ranks between #266 and #1,009 across 5 built-in weightings — best under "Coding-focused", worst under "All sources".
Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →