Google Cloud: LLM prices, GPUs and regions
Google Cloud sells 9 of the compared models per token and prices 5 kinds of GPU for hosting your own, across 2 regions.
Models sold per token
| Model | Input, per 1M tokens | Output, per 1M tokens | Cheapest region |
|---|---|---|---|
| DeepSeek V3.1 | $0.60 | $1.70 | global |
| Gemini 2.5 Flash | $0.30 | $2.50 | global |
| Gemini 2.5 Flash Lite | $0.10 | $0.40 | global |
| Gemini 2.5 Pro | $1.25 | $10.00 | global |
| Llama 3.3 70B | $0.72 | $0.72 | global |
| Llama 4 Maverick 17B | $0.35 | $1.15 | global |
| Llama 4 Scout 17B | $0.25 | $0.70 | global |
| OpenAI gpt-oss 120B | $0.09 | $0.36 | global |
| OpenAI gpt-oss 20B | $0.07 | $0.25 | global |
GPUs for hosting your own
| GPU | From, per GPU-hour | Cheapest instance | GPUs in it | Region |
|---|---|---|---|---|
| L4 | $0.7068 | g2-standard-4 | 1 × 24 GB | us-central1 |
| A100 | $3.48 | a2-megagpu-16g | 16 × 40 GB | us-central1 |
| RTX PRO 6000 | $4.50 | g4-standard-48 | 1 × 96 GB | us-central1 |
| H200 | $10.60 | a3-ultragpu-8g | 8 × 141 GB | us-central1 |
| H100 | $11.06 | a3-highgpu-8g | 8 × 80 GB | us-central1 |
Regions with prices
global, us-central1
List prices collected 5 October 2026, refreshed every three hours. No negotiated rates or taxes.