Google: Gemma 4 31B

google · Released Apr 2, 2026
DFO score #631 / 749
6.5 leaderboard score
Speed #331 / 355
35.6 tok/s
Input Price #360 / 749
$0.200 per 1M tokens
Output Price #295 / 749
$0.400 per 1M tokens
Context #208 / 749
262,144 tokens

Analysis Summary

Google's Gemma 4 31B pairs a 262K context window with image and video input, tool use, and function calling. Its coding and instruction-following measurements are relatively stronger than its overall intelligence index, and it has a stronger coding profile than Gemma 4 26B A4B in this batch. Listed rates are $0.09 per million input tokens and $0.34 per million output tokens.

This profile can serve document-heavy content operations, visual analysis, and coding assistance where a long context is useful. The weak agentic index is a material constraint for autonomous workflows, so teams should validate tool execution and retain approval steps rather than delegate complex tasks unattended. Its multimodal inputs are useful for work involving visual assets, but the supplied data does not establish a broad advantage in reasoning.

Adopt it as a practical multimodal assistant for supervised tasks, not as the default for high-stakes autonomous agents.

Assessed September 23, 2026

Editorial notes

Gemma 4 31B offers a 262K context, image and video input, and tool support, with stronger coding and instruction-following measurements than its agentic result suggests.

The DFO score ranks models on the Artificial Analysis Intelligence Index and is refreshed as new models launch. The write-up and verdict are our own. Feedback?

DFO Verdict

Gemma 4 31B offers a 262K context, image and video input, and tool support, with stronger coding and instruction-following measurements than its agentic result suggests.

#631 of 749 overall Down 67 this week

Benchmarks

Intelligence Index 4.4 #338 of 454 · Artificial Analysis
Coding Index 43.4 #38 of 199 · Artificial Analysis
Agentic Index 4.2 #116 of 181 · Artificial Analysis
Creative Writing 1,419 #26 of 32 · Arena rating
Instruction Following 1,454 #17 of 35 · Arena rating
GPQA Diamond 85.7%
HLE 22.7%
SciCode 43.4%
TerminalBench Hard 36.4%
τ²-Bench 59.9%
IFBench 75.6%
LCR 62%

Artificial Analysis data refreshed Nov 16, 2026. Arena ratings from the Nov 9, 2026 text leaderboard, with style control. Bars: magenta = reasoning, ink = coding and agents, cyan = instructions and long context.

How Google: Gemma 4 31B compares

Google: Gemma 4 31B ranks #338 of 454 AI models we track for overall intelligence, #38 of 199 for coding, #116 of 181 for agentic tasks. Its 262K-token context window is larger than 72% of the models we list. At $0.20 per million input tokens it is cheaper than 52% of comparable models.

Position in the field
DFO score: ahead of 16% of models #631
Speed: faster than 7% of models #331
Price: cheaper than 52% of models #360
Context: larger than 72% of models #208
worst in fieldmedianbest in field
Price vs frontier peers · $ per 1M tokens
Google: Gemma 4 31B $0.20 in $0.40 out
Anthropic: Claude Fable 5.2 $8.00 in $40.00 out
Google: Gemini 4 Carbon $2.00 in $10.00 out
Google: Gemini 4 Argon $2.00 in $10.00 out

Dark bar = input · light bar = output, scaled to the priciest peer.

Context window vs peers · tokens

1M tokens ≈ 8 full-length novels or ~2,500 pages of business documents in a single request.

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.20 $0.000200
Output $0.40 $0.000400

What would Google: Gemma 4 31B cost your business?

Pick the job that looks most like yours, then fine-tune with the sliders. Estimates update live.

A website chatbot handling around 100 customer conversations a day, a few short messages each.

3,000
One request is one message, email, draft or automation call.
1,200 tokens

$0.936/mo Google: Gemma 4 31B $0.0003 per request
$63.36/mo Anthropic: Claude Fable 5.2 $0.02 per request
$15.84/mo Google: Gemini 4 Carbon · best value $0.0053 per request

Full calculator with 749 models → Price Calculator

DFO AI AUTOMATION

These numbers get smaller with the right architecture.

We route routine calls to cheap models and save Google: Gemma 4 31B for the hard ones. Most clients cut their estimate by 60-80%.

Talk to our team

About Google: Gemma 4 31B

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function..

Frequently asked questions about Google: Gemma 4 31B

How much does Google: Gemma 4 31B cost?

Google: Gemma 4 31B costs $0.20 per million input tokens and $0.40 per million output tokens.

What is the context window of Google: Gemma 4 31B?

Google: Gemma 4 31B has a context window of 262,144 tokens (262K).

Is Google: Gemma 4 31B good for coding?

On our coding benchmark index, Google: Gemma 4 31B ranks #38 of 199 models, placing it in the top quartile of the field for code generation and debugging.

What can Google: Gemma 4 31B do?

Google: Gemma 4 31B supports image/vision input, tool use, and function calling.

Who created Google: Gemma 4 31B?

Google: Gemma 4 31B is developed by Google and was released on April 2, 2026.

© 2026 Design for Online Ltd. Registered in England and Wales No. 10328553. VAT Registered. Design for Online® and Forerunner® are registered trademarks.