Tested on 16 benchmarks · BenchGecko score 33.3. Top scores: OpenBookQA (71.5%), HellaSwag (61.9%), PIQA (54.6%).
BIG-Bench Hard. 23 challenging tasks from BIG-Bench where prior language models fell below average human performance.
HuggingFace MuSR (Multi-Step Reasoning). Tests multi-hop reasoning requiring chaining multiple facts together.
Grade school math word problems. 8,500 problems testing multi-step arithmetic reasoning. A foundational math benchmark.
HuggingFace evaluation of MATH Level 5 problems. Competition math requiring advanced reasoning and proof construction.
Elementary science questions with access to a small book of core science facts. Tests reasoning beyond memorization.
Sentence completion requiring commonsense reasoning about physical and social situations. Tests real-world understanding.
Physical Intuition QA. Tests understanding of everyday physical interactions and commonsense physics.
- Typetext
- ContextN/A
- ReleasedJan 2024
- LicenseOpen Source
- Statusbenchmark-only
Frequently Asked Questions
Key facts · as of 2026-04-09
- Gemma 2B by Google DeepMind. BenchGecko score 33.3, rank 233 of 312 scored models (normalized average of public benchmark scores).
- List price n/a input · n/a output per 1M tokens (as of 2026-04-09).
How to cite · data as of 2026-04-09
Gemma 2B · benchmarks, pricing and providers. BenchGecko, data as of 2026-04-09. https://benchgecko.ai/model/gemma-2b
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