Gemma 4 31B
As of Nov 7, 2026, the cheapest Gemma 4 31B API price is $0.1 per 1M input tokens (CoreWeave) and $0.34 per 1M output tokens (CoreWeave). Gemma 4 31B is offered by 5 providers in the gemma family, open weights, reasoning enabled. 2 providers list it as a free-tier offer.
API pricing by provider
Metered per-token prices. Cheapest input and output are highlighted; † marks a context-tier surcharge. How prices are ranked. Seeing a wide range? Why the same LLM costs different amounts.
| CoreWeaveopengoogle/gemma-4-31B-it | Cheapest$0.1 | Cheapest$0.34 | $0.1 | — | 262k | — |
| NanoGPTopengoogle/gemma-4-31b-it | $0.14 | $0.4 | $0.05 | — | 262k | 19d agoout −11.1% |
| Neuralwattopengemma-4-31b | $0.144 | $0.42 | $0.0144 | — | 262k | — |
| NanoGPTopenTEE/gemma4-31b | $0.4 | $1 | $0.4 | — | 262k | — |
| OpenRouterfreegoogle/gemma-4-31b-it:free | free | free | — | — | 262k | — |
| Poefreegoogle/gemma-4-31b | free | free | — | — | 262k | — |
Gemma 4 31B price history
Output price per 1M tokens since tracking began on Sep 10, 2026, for the one provider whose rate moved. Each step is a recorded change. All recent price changes.
| Date | Provider | 1M input | 1M output | Change |
|---|---|---|---|---|
| NanoGPT | $0.1 → $0.14 | $0.45 → $0.4 | −11.1% |
Gemma 4 31B pricing FAQ
What is the cheapest Gemma 4 31B API?
As of Nov 7, 2026, CoreWeave has the lowest Gemma 4 31B output price at $0.34 per 1M tokens, and CoreWeave has the lowest input price at $0.1 per 1M tokens. Prices are metered per-token rates in USD; free tiers and subscription plans are excluded.
How many providers offer Gemma 4 31B?
5 providers list Gemma 4 31B on Sovyron; 4 of them sell it at a metered per-token price.
Is Gemma 4 31B free?
2 providers list a free-tier offer: OpenRouter, Poe. Free tiers usually have rate limits.
What is the context window of Gemma 4 31B?
Gemma 4 31B supports a 262k-token context window and up to 262k output tokens.
Token cost calculator
What a month costs at one provider's published rates. Enter the tokens you expect to send and receive; cached tokens use the cached-input price where the provider lists one, and count as ordinary input where it does not.
Estimated
—