Gemma 4 31B pricing & benchmarks
Gemma 4 31B is available from Google at $0.100 per million input tokens and $0.340 per million output tokens ($0.160 blended at 3:1). It accepts up to 262,144 tokens of context. 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...
Pricing
Per million tokens · OpenRouter
Input$0.100
Output$0.340
Blended (3:1)$0.160
Specification
google/gemma-4-31b-it
Context window262K
Max output262K
ReleasedApr 2026
LicenceOpen weights
Modalitiesin:text, in:image, in:video, out:text
Benchmarks
Independently run — not vendor self-reports. Every score links its source.
| Benchmark | Score | Config | Run | Source |
|---|---|---|---|---|
| GPQA Diamond | 75.8% ±3.1 | minimal effort | 2026-08-06 | Independently run · Epoch AI |
| SimpleQA Verified | 9.6% ±0.9 | — | 2026-04-04 | Independently run · Epoch AI |
| OTIS Mock AIME | 73.3% ±6.7 | minimal effort | 2026-08-06 | Independently run · Epoch AI |
Scores published by Epoch AI, 'AI Benchmarking Hub', used under CC BY 4.0.
Cheaper models that score at least as well on GPQA Diamond
Strictly lower blended price AND an equal-or-higher GPQA Diamond score. One benchmark is one dimension — a model that wins here may still be weaker at your specific task.
| Model | Provider | Blended $/1M | GPQA Diamond | |
|---|---|---|---|---|
| gpt-oss-120b | OpenAI | $0.070 (−56%) | 75.8% | Compare → |
| DeepSeek V4 Flash 0731 | DeepSeek | $0.113 (−30%) | 91.0% | Compare → |
Where to run Gemma 4 31B
18 options across 16 providers · 7.6× between cheapest and dearest. Blended at 3:1. ⚠ marks an option serving less context than the best available, or re-pricing above a prompt-length threshold. Bracketed labels are the provider's own service tier or region.
| Host | Blended $/1M | In / Out | Context | Weights | Uptime 24h |
|---|---|---|---|---|---|
| OpenInference | $0.147 | $0.080 / $0.350 | 262K | bf16 | 92.8% |
| DeepInfra (turbo) | $0.152 | $0.090 / $0.340 | 262K | fp4 | 99.9% |
| CoreWeave | $0.160 | $0.100 / $0.340 | 262K | bf16 | 96.2% |
| Venice | $0.180 | $0.120 / $0.360 | 256K | bf16 | 99.7% |
| Chutes ⚠ 131K context, not 262K |
$0.182 | $0.120 / $0.370 | 131K | fp4 | 92.1% |
| DeepInfra | $0.193 | $0.130 / $0.380 | 262K | fp8 | 93.3% |
| SiliconFlow | $0.198 | $0.130 / $0.400 | 262K | fp8 | 91.4% |
| Novita | $0.205 | $0.140 / $0.400 | 262K | bf16 | 95.2% |
| Friendli | $0.205 | $0.140 / $0.400 | 262K | — | 100.0% |
| Morph ⚠ 175K context, not 262K |
$0.205 | $0.140 / $0.400 | 175K | fp4 | 95.8% |
| Crusoe | $0.205 | $0.140 / $0.400 | 262K | — | 98.8% |
| Parasail | $0.212 | $0.150 / $0.400 | 262K | fp8 | 96.8% |
| Phala | $0.227 | $0.150 / $0.460 | 262K | — | 93.5% |
| ModelRun | $0.302 | $0.220 / $0.550 | 262K | fp4 | 96.7% |
| Together | $0.425 | $0.280 / $0.860 | 262K | — | 85.3% |
| Together | $0.535 | $0.390 / $0.970 | 262K | — | 95.8% |
| SambaNova ⚠ 131K context, not 262K |
$0.573 | $0.380 / $1.15 | 131K | — | 99.7% |
| Cerebras ⚠ 131K context, not 262K |
$1.11 | $0.990 / $1.49 | 131K | fp16 | 98.8% |
Host prices and uptime from OpenRouter, synced 2026-08-08. Lower-precision weights (fp4, fp8) can cost less and score worse than the same model at bf16 — the score above is not host-specific.
Compare Gemma 4 31B
FAQ
How much does Gemma 4 31B cost?
Gemma 4 31B costs $0.100 per million input tokens and $0.340 per million output tokens on OpenRouter. At a 3:1 input:output mix that blends to $0.160 per million tokens.
What is Gemma 4 31B's context window?
Gemma 4 31B accepts up to 262,144 tokens of context, and can generate up to 262,144 output tokens.
How good is Gemma 4 31B on benchmarks?
Gemma 4 31B scores 75.8% on GPQA Diamond, independently run and published by Epoch AI. 3 benchmark results are listed on this page, each with its run date and source.
Is there a cheaper model as good as Gemma 4 31B?
Yes — 2 models in our catalogue cost less per token than Gemma 4 31B and score at least as high on the same benchmark. The cheapest is gpt-oss-120b at $0.070 per million tokens blended.