LivingPaper

Settings

The language models LivingPaper talks to. You bring your own: any provider with an OpenAI-compatible chat API, such as OpenAI or a local server like Ollama or LM Studio.

Language models

  • deepseek 4.1

    For analysis For chat

    Address
    https://openrouter.ai/api/v1
    Model
    deepseek/deepseek-v4.1-flash
    Key
    Saved, encrypted
    Delete

    This deletes the model and its saved key. It cannot be undone.

Add a language model

Type the name, the address and the key, then get the models your provider offers. That request also checks the key.

Only for you, to tell your models apart. For example OpenAI or My local model.

The address of the API, for example https://api.openai.com/v1

Kept encrypted on this computer and never shown again. Leave it empty for a local server that needs none.

The provider is asked for its models: this takes a few seconds. The page waits for the answer.

Which model does what

One model can do both. A cheaper one can read books while a better one chats.