Please add SiliconFlow as a supported AI Connector in Remalt.
SiliconFlow provides one OpenAI-compatible API for 200+ AI models across LLMs, vision, image, video and audio. It supports models from providers such as DeepSeek, Qwen, Kimi, MiniMax, GLM, OpenAI open-weight models and others.
This would give Remalt users a much broader and more cost-efficient choice of models without needing separate integrations for every provider.
Useful capabilities for the connector would include:
API key authentication
OpenAI-compatible API support
Dynamic model selection
Chat and reasoning models
Vision models
Image generation
Video generation
Audio models
Streaming
Function/tool calling where supported
Embeddings and reranking
Usage and cost tracking
Secure API-key storage
SiliconFlow also offers pay-as-you-go serverless inference with no minimum spend and very competitive pricing across many models.
Because the API is OpenAI-compatible, the implementation may also be relatively straightforward compared with building a completely new provider interface. SiliconFlow explicitly describes its service as providing one API across models with OpenAI compatibility.
Official documentation:
SiliconFlow Documentation
Pricing:
SiliconFlow Pricing
