Home / Guides / Apps and workflows
Connect DeepSeek and GLM to Dify
Updated 2026-10-18 · Use the OpenAI-API-compatible model provider to bring DeepSeek, GLM and other models into Dify apps and workflows.
Connection details
- Base URL (OpenAI compatible)
- https://api.aifans.fan/v1
- API key
- Create one in the console; it starts with sk- (no account yet? sign up)
- Recommended models
- deepseek-v4.1-flash, deepseek-v4-pro (all models and prices)
Dify is a popular open-source platform for building LLM apps: knowledge-base Q&A, agents and workflows. Its “OpenAI-API-compatible” model provider connects every chat model on AIFans API.
Setup
- Avatar (top right) → Settings → Model Provider, and install OpenAI-API-compatible from the plugin marketplace (recent Dify versions ship model providers as plugins).
- Under that provider click “Add Model”: model type LLM, model name deepseek-v4.1-flash.
- API key your key; API endpoint URL https://api.aifans.fan/v1.
- Context size can be 128000; set max tokens as you need (e.g. 8192); Function calling “supported”; Vision “supported” for models that accept images.
- Save, then choose the model in your app's model settings. For harder reasoning, add deepseek-v4-pro the same way.
Notes
- A self-hosted Dify must be able to reach our public HTTPS address.
- We do not offer embedding or rerank models; pick those from another provider in Dify.
- Workflows that call the model in bulk count against each key's per-minute request limit (100 by default); give a busy workflow its own key.