Gemma 3 27B

Facts

Released

2025-03-01 the day the repository was first published on Hugging Face Hugging Face, read 2026-09-24

Licence

gemma the licence the model card declares Hugging Face, read 2026-09-24

Open weights

Yes the weights are published in this Hugging Face repository, behind the maker's access form Hugging Face, read 2026-09-24

Parameters

27.4 billion counted from the safetensors weight files Hugging Face, read 2026-09-24

Active parameters

Unknown

Knowledge cutoff

2024-08-31 the knowledge cutoff OpenRouter lists OpenRouter, read 2026-09-24

Context window

• 131,072 tokens the maximum position embeddings in the model configuration model configuration, read 2026-09-24
• 131,072 tokens the context length OpenRouter lists OpenRouter, read 2026-09-24

Longest output

• 117,964 tokens the largest output OpenRouter's first provider allows OpenRouter, read 2026-09-24

Tool calling

• No the chat template in the tokenizer configuration has no place for tool definitions Hugging Face, read 2026-09-24
• Yes OpenRouter lists tools among the supported parameters OpenRouter, read 2026-09-24

Structured output

• Yes OpenRouter lists structured outputs among the supported parameters OpenRouter, read 2026-09-24

Reasoning controls

Unknown

Inputs and outputs

text, image in, text out Hugging Face, read 2026-09-24

Good for

• Vision: Hugging Face files it under image-text-to-text Hugging Face, read 2026-09-24

Prices

Run it on your own hardware

Weights are the sizes of the files a source lists, or an estimate from the parameter count where none does. Memory is an estimate: weights plus KV cache plus 512 MiB and 5 percent of the weights for runtime buffers. Check it against your hardware.

Start it with a local runtime

• Ollama

  ollama run hf.co/MaziyarPanahi/gemma-3-27b-it-GGUF:Q4_K_S

• LM Studio

  lms get MaziyarPanahi/gemma-3-27b-it-GGUF
  lms server start

• llama.cpp server

  llama-server -hf MaziyarPanahi/gemma-3-27b-it-GGUF:Q4_K_S --jinja

• vLLM

  vllm serve google/gemma-3-27b-it

• SGLang

  sglang serve --model-path google/gemma-3-27b-it --port 30000

Use it from your harness

Set up for DeepInfra with the model google/gemma-3-27b-it. Each endpoint page has the same setup for its own address.

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "deepinfra": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "DeepInfra",
      "options": {
        "baseURL": "https://api.deepinfra.com/v1/openai",
        "apiKey": "{env:DEEPINFRA_TOKEN}"
      },
      "models": {
        "google/gemma-3-27b-it": {
          "name": "Gemma 3 27B",
          "limit": {
            "context": 131072,
            "output": 117964
          }
        }
      }
    }
  }
}

{
  "providers": {
    "deepinfra": {
      "baseUrl": "https://api.deepinfra.com/v1/openai",
      "api": "openai-completions",
      "apiKey": "$DEEPINFRA_TOKEN",
      "models": [
        {
          "id": "google/gemma-3-27b-it"
        }
      ]
    }
  }
}

export ANTHROPIC_BASE_URL="https://api.deepinfra.com/anthropic"
export ANTHROPIC_AUTH_TOKEN="$DEEPINFRA_TOKEN"
export ANTHROPIC_MODEL="google/gemma-3-27b-it"
claude

Other ways to reach it

• Hugging Face, model google/gemma-3-27b-it
• Kilo Gateway, model google/gemma-3-27b-it
• Nebius Token Factory, model google/gemma-3-27b-it
• NovitaAI, model google/gemma-3-27b-it
• OpenRouter, model google/gemma-3-27b-it
• STACKIT, model google/gemma-3-27b-it
• Ollama, model hf.co/MaziyarPanahi/gemma-3-27b-it-GGUF:Q4_K_S

Published results

These are results other people published. Baltor did not run them and does not rank models by them.

• Results the maker published on the model card published by google Hugging Face, read 2026-09-24
• Artificial Analysis Intelligence Index: 4.9 published by Artificial Analysis OpenRouter, read 2026-09-24
• Artificial Analysis Coding Index: 10.1 published by Artificial Analysis OpenRouter, read 2026-09-24
• Artificial Analysis Agentic Index: 0.1 published by Artificial Analysis OpenRouter, read 2026-09-24