Product updates
LLM gateway API now open for enquiries
Our LLM gateway API is now open for enquiries. Through one OpenAI-compatible interface, businesses can reach several leading models and mana…
New GPU compute servers available
To meet demand for model training and inference, we now offer GPU compute servers in 1- to 8-GPU configurations with CUDA, PyTorch and other…
RAG knowledge base supports more file formats
The RAG knowledge base now parses more formats, so you can import PDF, Word, Excel and web content directly.…
Kubernetes service adds autoscaling
The Kubernetes service now autoscales on CPU, memory and custom metrics — growing at peak times and shrinking when quiet to save cost.…