ScaledThought
← All models

Google DeepMind

Gemma 4 31B

Live · US + CAtextimageaudioApache 2.0

The largest Gemma 4 model. The family moved from Google's custom license to full Apache 2.0.

Price per 1M tokens

Input

$0.09

Cached input

$0.05

Output

$0.34

Priced per model · Same in US and CA · Free tier included · Pricing

AA Index

19

#13 of 14
open models we serve

Specs

Model id
gemma-4-31b
Weights released
Apr 2, 2026
Parameters
31B
Architecture
Dense
Context
256K tokens (262,144)
Input
text, image, audio
License
Apache 2.0
Regions
us, ca

Benchmarks

  • Artificial Analysis Intelligence Index19

Sources: blog.google · huggingface.co · artificialanalysis.ai

Use it

Three lines of code, or zero.

OpenAI-compatible. Use the model id below, or auto to let the router decide when this model is worth it.

Quickstart→
import os
from openai import OpenAI

client = OpenAI(base_url=os.environ["ST_BASE_URL"], api_key=os.environ["ST_API_KEY"])
resp = client.chat.completions.create(
    model="gemma-4-31b",
    messages=[{"role": "user", "content": "Hello"}],
)

Opens your agent with the prompt typed. You press Enter.