Model passport
Vicuna 13b
vicuna-13b- Open weights
- Llama 2 Community licence
- Released 2023-03-30(curated)
- Overall · LMArena
- 1,142#362 of 396
- Best percentile
- 8.8Overall
- Cohort standing
- 26.1median 24.9
What the data says
Benchmarked on 4 of 27 tracked axes. Every figure below is the model's own published result on that board.
Weakest axes
- 7.3 pct — Hard prompts (1,132)
- 7.7 pct — Coding (1,164)
- 8.8 pct — Instruction following (1,117)
Uneven: 1.5 percentile points between its best and worst axis — 8.8 pct on Overall versus 7.3 pct on Hard prompts.
- Overall8.8 pct1,142Verify →
- Instruction following8.8 pct1,117Verify →
- Coding7.7 pct1,164Verify →
- Hard prompts7.3 pct1,132Verify →
- WebDev ArenaNot benchmarked
- Vision ArenaNot benchmarked
- Factuality ArenaNot benchmarked
- Search ArenaNot benchmarked
- Document ArenaNot benchmarked
- Agent ArenaNot benchmarked
- GPQA DiamondNot benchmarked
- AIME (math)Not benchmarked
- FrontierMathNot benchmarked
- SimpleQA VerifiedNot benchmarked
- ARC-AGI-2Not benchmarked
- Humanity's Last ExamNot benchmarked
- SimpleBenchNot benchmarked
- LiveBenchNot benchmarked
- BFCL v4 (function calling)Not benchmarked
- SWE-bench Verified (live board)Not benchmarked
- Terminal-Bench 4.0Not benchmarked
- MMLU-Pro (Artificial Analysis)Not benchmarked
- LiveCodeBench (Artificial Analysis)Not 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.
vicuna-13b stands at 26.1 (raw mean 8.2) — measured on 4 of 27 axes.
vicuna-13b stands at 26.1 versus a cohort median of 24.9 across 28 models released within 180 days — above the peer median.
vicuna-13b scores at percentile 7.3 on Hard prompts.
vicuna-13b scores at percentile 7.7 on Coding.
vicuna-13b ranks between #377 and #793 across 5 built-in weightings — best under "Conversation-first", 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 →