Model Garden
Use open models like Llama, DeepSeek, Qwen, Gemma and gpt-oss through Google's Vertex AI, as a managed API or your own endpoint.
Overview
Vertex AI Model Garden is Google Cloud's catalogue of AI models. Many open-weight models, including Llama, DeepSeek, Qwen and gpt-oss, are available as fully managed APIs billed per million tokens, with no servers to run. Others, including Google's Gemma family, deploy in a few clicks to a dedicated Vertex AI endpoint billed per hour of the machine it runs on.
Specifications
| Hosting type | Serverless API |
|---|---|
| Model families | Llama, Mistral, DeepSeek, Qwen, gpt-oss, Gemma |
| LLM models | Llama 3.3 70B Instruct, Mistral Small 3.1, DeepSeek-V3.1, DeepSeek-R1, Qwen3 235B-A22B, gpt-oss-120b, gpt-oss-20b, Gemma 2 9B, Gemma 2 27B, Gemma 3 4B, Gemma 3 12B, Gemma 3 27B |
| Pricing range | US$0.07 – US$5.4 |
| Pricing model | Pay per token, Per GPU-hour |
| Free tier available | No |
| Data center locations | United States, Canada, Brazil, Chile, Mexico, Belgium, Netherlands, Germany, Finland, United Kingdom, France, Italy, Spain, Switzerland, Poland, Sweden, Israel, Qatar, Saudi Arabia, South Africa, India, Japan, South Korea, Taiwan, Hong Kong SAR China, Singapore, Indonesia, Australia |
| Hardware / GPU options | Managed APIs run on Google's infrastructure. Self-deployed models run on NVIDIA T4, L4, A100, H100, H200, B200 and GB200 GPUs, plus Google Cloud TPUs (v5e, v6e Trillium and later). |
| VRAM & memory | |
| Maximum model size | 671B |
| OpenAI-compatible API | Yes |
News about Google Cloud
All news →LLM Hosting Roundup: Anthropic’s capacity crunch, Vera Rubin and DeepSeek’s $15B (23 Oct 2026)
Anthropic's hunt for cloud capacity, NVIDIA's DSX POD and new Vera Rubin inference benchmarks, Dell's first Groq 3 rack, DeepSeek's reported $15B raise and Reflection's open-weights Beam model.
