Model specification · Google

Gemma 4 31B

google/gemma-4-31b-it

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...

SpecificationsReported values
Context window
262K
Maximum output
16K
Architecture
text+image+video->text
Tokenizer
Gemma
Knowledge cutoff
—
Moderated
No
Input modalities
image text video
Output modalities
text
Full pricingUSD where reported
PriceStandardfree
Input$0.090 / Mfree / M
Output$0.340 / Mfree / M
Benchmark indexesReported indexes
Intelligence
15.4
Coding
43.4
Agentic
6.7
Design Arena Elo
—
Example workload costsStandard pricing only
WorkloadInput / outputCache shareEstimated cost
Quick check10K / 1K × 10%$0.00124
Long analysis100K / 4K × 150%$0.00836
Workday20K / 2K × 1,00030%$2.24
More from GoogleAll Google models
ModelContextInput / MOutput / MIntel.
Gemini 3.8 Flash1.0M$0.750$3.7541.2
Gemini 3.7 Flash1.0M$0.750$3.7539.4
Gemini 3.6 Flash1.0M$0.750$3.7534.3
Gemini 3.5 Flash1.0M$1.50$9.0033
Gemini 3.1 Pro Preview1.0M$2.00$12.0030.4
Gemini 3.5 Flash Lite1.0M$0.300$2.5022.7
Gemini 2.5 Pro1.0M$1.25$10.0016.7
Gemini 3.1 Flash Lite Preview1.0M$0.250$1.5016
Gemma 3 27B131K$0.080$0.4504.9
Gemma 3 12B131K$0.050$0.1503.8