---
name: Gemma 4 31B Instruct
slug: gemma-4-31b
type: Multimodal
providers:
  - togetherai
  - poe
  - nano-gpt
  - venice
  - berget
  - neuralwatt
  - wandb
  - novita-ai
  - chutes
  - openrouter
description: Open Gemma instruction model for efficient chat and self-hosted deployments
tags:
  - open-source
  - reasoning
  - tool-use
  - multimodal
  - gemma
is_open_source: true
context_window: 262144
max_output: 262144
pricing:
  - image: 0
    input: 0.275
    output: 0.55
    provider: berget
  - image: 0
    input: 0.12
    output: 0.37
    provider: chutes (TEE)
  - image: 0
    input: 0.4
    output: 1
    provider: nano-gpt
  - image: 0
    input: 0.4
    output: 1
    provider: nano-gpt (TEE)
  - image: 0
    input: 0.144
    output: 0.42
    provider: neuralwatt
  - image: 0
    input: 0.14
    output: 0.4
    provider: novita-ai
  - image: 0
    input: 0
    output: 0
    provider: openrouter (Free Tier)
  - image: 0
    input: 0
    output: 0
    provider: poe
  - image: 0
    input: 0.39
    output: 0.97
    provider: togetherai
  - image: 0
    input: 0.12
    output: 0.36
    provider: venice
  - image: 0
    input: 0.1
    output: 0.34
    provider: wandb
release_date: '2026-04-02'
capabilities:
  audio: true
  vision: true
  tool_use: true
  citations: false
  reasoning: true
---
Gemma 4 31B Instruct is a Multimodal model.
Open Gemma instruction model for efficient chat and self-hosted deployments