Google models: prices and the hardware each needs
26 models, 5 sold per token by 7 vendors. Vendors are in alphabetical order; the price is the cheapest list price among them, per million tokens.
| Model | Size | Sold per token by | From, input | Output | GPU memory |
|---|---|---|---|---|---|
| codegemma-7b-it | 8.54 B | not sold per token | 21.3 GB | ||
| Gemini 2.5 Flash | closed | DeepInfra, Gemini API, Snowflake | $0.30 | $2.50 | |
| Gemini 2.5 Flash Lite | closed | Databricks, Gemini API, Snowflake | $0.10 | $0.40 | |
| Gemini 2.5 Pro | closed | DeepInfra, Gemini API | $1.25 | $10.00 | |
| Gemma 3 27B | 27.4 B | DeepInfra, Nebius | $0.08 | $0.16 | 68.6 GB |
| Gemma 4 26B A4B | 25.8 B | Cloudflare, DeepInfra, Snowflake, Telnyx | $0.07 | $0.34 | 64.5 GB |
| gemma-1.1-2b-it | 2.51 B | not sold per token | 6.3 GB | ||
| gemma-2-27b-it | 27.23 B | not sold per token | 68.1 GB | ||
| gemma-2-2b | 2.61 B | not sold per token | 13 GB | ||
| gemma-2-2b-it (google) | 2.61 B | not sold per token | 6.5 GB | ||
| gemma-2-9b | 9.24 B | not sold per token | 46.2 GB | ||
| gemma-2-9b-it | 9.24 B | not sold per token | 23.1 GB | ||
| gemma-2b | 2.51 B | not sold per token | 6.3 GB | ||
| gemma-2b-it (google) | 2.51 B | not sold per token | 6.3 GB | ||
| gemma-3-12b-it | 12.19 B | not sold per token | 30.5 GB | ||
| gemma-3-1b-it (google) | 1 B | not sold per token | 2.5 GB | ||
| gemma-3-1b-pt | 1 B | not sold per token | 2.5 GB | ||
| gemma-3-270m-it | 0.27 B | not sold per token | 0.7 GB | ||
| gemma-3-4b-it | 4.3 B | not sold per token | 10.8 GB | ||
| gemma-4-12B-it | 11.96 B | not sold per token | 29.9 GB | ||
| gemma-4-26B-A4B | 26.54 B | not sold per token | 66.3 GB | ||
| gemma-4-31b-it | 31.27 B | not sold per token | 78.2 GB | ||
| gemma-7b | 8.54 B | not sold per token | 21.3 GB | ||
| gemma-7b-it | 8.54 B | not sold per token | 21.3 GB | ||
| medgemma-27b-text-it | 27.01 B | not sold per token | 67.5 GB | ||
| shieldgemma-9b | 9.24 B | not sold per token | 23.1 GB |