Introduction
Welcome to Bitdeer AI — serverless inference APIs for self-deployed text and vision chat models.
Welcome to Bitdeer AI
Bitdeer AI provides serverless HTTP APIs for self-deployed text and vision chat models. Deploy nothing — use https://api-inference.bitdeer.ai/v1 and your API key.
What Bitdeer AI offers
| Capability | Models | Use case |
|---|---|---|
| Text chat | zai-org/GLM-5.2, moonshotai/Kimi-K2.6, Qwen/Qwen3.5-397B-A17B, MiniMaxAI/MiniMax-M2.5, nvidia/Nemotron-3-Nano-Omni-30B-A3B, nvidia/NVIDIA-Nemotron-3-Super-120B-A12B, deepseek-ai/DeepSeek-V4-Pro, XiaomiMiMo/MiMo-V2.5-Pro, nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B | Chatbots, code gen, summarization, translation |
| Vision | moonshotai/Kimi-K2.6, Qwen/Qwen3.5-397B-A17B, nvidia/Nemotron-3-Nano-Omni-30B-A3B | Image understanding in messages |
| Audio | nvidia/Nemotron-3-Nano-Omni-30B-A3B | Audio input via input_audio |
Base URL
All API requests go to:
Calling the API
Endpoints follow common chat-completions JSON shapes (POST /v1/chat/completions, etc.). Use cURL, your own HTTP stack, or any client that lets you set a custom base URL. Step-by-step requests and Python / JavaScript snippets are in Get Started and SDKs.
How this documentation is organized
| Section | What it covers |
|---|---|
| Get Started | Get your API key and send a first request (cURL or client libraries) |
| LLM | Streaming, Vision, Function Calling, Structured Outputs |
| Integrations | Cursor, LangChain, OpenAI Agents SDK, Dify, LiteLLM |
| Language clients | Python and JavaScript (openai client + Bitdeer base_url) |
| Examples | Complete runnable examples |
| API Reference | Endpoint-level documentation with schemas |
| GPU Cloud | Dedicated GPU instances, networking, and Instance API reference |
Quick start
Get an API key
Sign up at bitdeer.ai and generate a key from Models → API Keys.
Install client libraries
Make your first request
See the complete walkthrough in Get Started → First Request.
Last updated on