Google

Gemma 4 31B

Released 2026-04-02

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...

You send / 1M

$0.200

tokens in

It replies / 1M

$0.400

tokens out

Memory

262K

tokens it can read at once

Typical request

Free

1,000 words in, 400 out

Data privacy

No training or retention: Yes. At least one provider serves it with zero data retention.

ProviderYou send / 1MIt replies / 1MUptimePrivacy policy
DeepInfra · fp8$0.160$0.320100%Read
CoreWeave · fp4$0.100$0.340100%Read
Venice · fp4$0.120$0.360100%Read
DekaLLM$0.140$0.40099%Read
Crusoe · bf16$0.140$0.40099%Read
DeepInfra · fp8$0.200$0.400100%Read
Parasail · fp8$0.150$0.400100%Read
Novita · bf16$0.140$0.40092%Read
DeepInfra · fp4$0.135$0.510100%Read
DeepInfra · fp8$0.300$0.600100%Read
ModelRun · fp4$0.750$1.00100%Read
SiliconFlow · fp8$0.750$1.0099%Read
Io Net$0.361$1.0999%Read
SambaNova$0.380$1.1595%Read

Based on OpenRouter and Vercel AI Gateway data. Confirm terms with the provider before sending sensitive data.

What it can do

  • Reads: image, text, video
  • Writes: text
  • Reasoning: Yes
  • Tool calling: Yes
  • Longest reply: 16K tokens
  • Intelligence score: 14.7 / 100 (Artificial Analysis)

Similar models

Appears in

Estimate your monthly cost with Gemma 4 31B →