Gemma 3 27B

Gemma 3 27B is a non-reasoning model from Google. 18 benchmarks count toward its score, in 8 categories.

availableShows if the model has enough results for an index.
IndexOverall score out of 100.27.0 ±3.8
CoverageShare of the index weight with results.100%
SpeedOutput tokens per second.31/s
Input / 1MUS dollars per 1M input tokens.Free
Output / 1MUS dollars per 1M output tokens.Free
ContextMaximum tokens in one request.32K
EloLMArena rating and rank.N/A

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

CapabilitiesScore per category, out of 100.

Out of 100
AgenticMulti-step tasks with tools.
26.4
CodingCode writing and repair.
25.6
ReasoningLogic problems and puzzles.
25.1
MultimodalTasks with images and text.
25.8
KnowledgeFacts and expert knowledge.
24.5
MultilingualTasks in many languages.
70.6
InstructionTasks with strict rules in the prompt.
21.4
MathMath problems.
29.6

Results

18 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.
MATH level 5Math74.0%35.1Epoch AI
EuroEval SwedishMultilingual58.4%75.1EuroEval
EuroEval SwedishMultilingual58.0%74.8EuroEval
EuroEval FrenchMultilingual58.0%74.7EuroEval
EuroEval PortugueseMultilingual57.7%74.4EuroEval
EuroEval FrenchMultilingual57.5%74.1EuroEval
EuroEval DutchMultilingual56.5%72.9EuroEval
EuroEval PolishMultilingual55.7%71.8EuroEval
EuroEval ItalianMultilingual53.8%69.4EuroEval
EuroEval PortugueseMultilingual53.5%69.1EuroEval
EuroEval PolishMultilingual51.8%67.0EuroEval
EuroEval GermanMultilingual50.7%65.6EuroEval
EuroEval SpanishMultilingual49.2%63.7EuroEval
Artificial Analysis MMMU-ProMultimodal48.0%25.8Artificial Analysis
GPQA diamondKnowledge47.7%22.3Epoch AI
EuroEval GermanMultilingual47.2%61.2EuroEval
EuroEval SpanishMultilingual43.8%57.0EuroEval
Artificial Analysis GPQA DiamondKnowledge42.8%12.4Artificial Analysis
Artificial Analysis IFBenchInstruction31.8%21.4Artificial Analysis
Artificial Analysis SciCodeCoding23.3%25.2Artificial Analysis
OTIS Mock AIME 2024-2025Math22.5%24.2Epoch AI
Artificial Analysis Omniscience AccuracyKnowledge13.0%30.0Artificial Analysis
τ²-Bench Tool-Agent-User EvaluationAgentic10.5%6.4Victor Barres et al.
Artificial Analysis Coding IndexCoding10.1%26.1Artificial Analysis
Artificial Analysis Long Context ReasoningReasoning7.3%13.3Artificial Analysis
Artificial Analysis Intelligence IndexKnowledge4.8%28.5Artificial Analysis
Artificial Analysis Humanity's Last ExamKnowledge4.4%29.3Artificial Analysis
Artificial Analysis Agentic IndexAgentic0.1%37.3Artificial Analysis
GDPval-AA normalizedAgentic0.0%35.6Artificial Analysis
Critical Physics TasksReasoning0.0%38.4Artificial Analysis
Chess PuzzlesReasoning0.0%23.5Epoch AI

18 benchmarks count, from 31 of 31 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 authorsEpoch AI, collected directlyCC BY — free to use and redistribute with attributionEuroEval, collected directlyMIT — the leaderboard site and its CSV routes are in the licensed repository

More from Google

Gemini 3.7 Flash72.3Gemini 3.8 Flash72.1Gemini 3.5 Flash67.3Gemini 3.6 Flash66.8Gemini 3.1 Pro64.7