Tested on 7 benchmarks · BenchGecko score 30.2. Top scores: PIQA (54.8%), HellaSwag (43.7%), Winogrande (41.6%).
BIG-Bench Hard. 23 challenging tasks from BIG-Bench where prior language models fell below average human performance.
Grade school math word problems. 8,500 problems testing multi-step arithmetic reasoning. A foundational math benchmark.
Physical Intuition QA. Tests understanding of everyday physical interactions and commonsense physics.
Sentence completion requiring commonsense reasoning about physical and social situations. Tests real-world understanding.
Commonsense coreference resolution. Tests understanding of pronoun references in ambiguous sentences.
- Typetext
- ContextN/A
- ReleasedJan 2024
- LicenseProprietary
- Statusbenchmark-only
Frequently Asked Questions
Key facts · as of 2026-06-22
- vicuna-13b-v1.1 by Unknown. BenchGecko score 30.2, rank 254 of 319 scored models (normalized average of public benchmark scores).
- List price n/a input · n/a output per 1M tokens (as of 2026-06-22).
How to cite · data as of 2026-06-22
vicuna-13b-v1.1 · benchmarks, pricing and providers. BenchGecko, data as of 2026-06-22. https://benchgecko.ai/model/vicuna-13b-v1-1
Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP