Skip to content
aiqbench

Google · model passport

Google: Gemma 3 27B

gemma-3-27b-it
  • Open weights
  • Gemma licence
  • 27.4B params
  • 131,072 context
  • Released 2025-03-12(curated)
11/27axes covered
Overall · LMArena
1,365#184 of 396
Best percentile
67.6HHEM Hallucination (Vectara)
Cohort standing
42.9median 45.8

What the data says

Benchmarked on 11 of 27 tracked axes. Every figure below is the model's own published result on that board.

Weakest axes

  • 12.2 pct — Factuality Arena (1,380)

Uneven: 55.4 percentile points between its best and worst axis — 67.6 pct on HHEM Hallucination (Vectara) versus 12.2 pct on Factuality Arena.

  • HHEM Hallucination (Vectara)67.6 pct92.6%Verify →
  • Overall53.8 pct1,365Verify →
  • Instruction following51 pct1,343Verify →
  • Hard prompts49.5 pct1,367Verify →
  • LiveBench40.5 pct48.4%Verify →
  • BFCL v4 (function calling)Dormant board40.2 pct29.5%Verify →
  • Coding38.4 pct1,358Verify →
  • Vision Arena35.3 pct1,160Verify →
  • GPQA Diamond31.7 pct47.7%Verify →
  • AIME (math)30 pct22.5%Verify →
  • Factuality Arena12.2 pct1,380Verify →
  • WebDev 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
  • 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
  • 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.

How to run it

Run it yourself

Repositorygoogle/gemma-3-27b-it
Parameters27.4B
ArchitectureGemma3ForConditionalGeneration
LicenceGemma
AccessGated — you must accept the repo's terms before downloading

Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "google/gemma-3-27b-it"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Ollama

We don't sync Ollama tags, so we won't print a command we can't verify — a wrong one would pull the wrong model. Search the Ollama library →

Unsloth fine-tuning recipes are model-specific and we don't derive them here. Unsloth docs →

Use it hosted

API model IDgoogle/gemma-3-27b-it
Context length131,072 tokens
Max output117,964 tokens
Acceptstext, image
Input price$0.08 per 1M tokens
Cached input price$0.04 per 1M tokens
Output price$0.35 per 1M tokens

API parameters this endpoint supports

Frequency penaltyLogit biasLogprobsMax tokensMin-pPresence penaltyRepetition penaltyResponse formatSeedStop sequencesStructured outputsTemperatureTool choiceTool callingTop-kTop logprobsTop-p

Example request

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"google/gemma-3-27b-it","messages":[{"role":"user","content":"Hello"}]}'

View on OpenRouter →

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.

  • gemma-3-27b-it stands at 42.9 (raw mean 40.9) — measured on 11 of 27 axes.

    standingtemplateVerify →

  • gemma-3-27b-it stands at 42.9 versus a cohort median of 45.8 across 148 models released within 180 days — below the peer median.

    cohorttemplateVerify →

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 →