Message Error: Model vscan-gpt-40 does not have an assigned provider

[Sean-Kelly]

Got this error after install when I tried to take a picture.
It is not clear how to fix this.

[RastislavKish]

Hello, thank you for opening an issue!

After the first launch of VScan, there are several steps you need to do in order to setup the providers you want to use, and also the models you want to use in your configs. Providers are services hosting or routing models i.e. OpenAI, Anthropic, Together.AI, etc. Because one model can be hosted by multiple services, for instance, gpt-4o can be accessed both through the official OpenAI server, or through a router like OpenRouter or Nano-GPT, VScan also requires that for each model you want to use you specify the provider you want to use. This can be even yourself, if you have a powerful-enough hardware and want to self-host vision models like Qwen 3 VL.

Please refer to the section Initial setup of the readme, which explains how to do the initial configuration.