UberLLM

Continue

Continue (the VS Code / JetBrains assistant) reaches UberLLM through its openai provider with a custom apiBase.

Setup

Edit your Continue config (~/.continue/config.yaml) and add a model:

name: My Config
version: 0.0.1
schema: v1

models:
  - name: UberLLM DeepSeek V3.1
    provider: openai
    model: deepseek/deepseek-v3.1
    apiBase: https://api.uberllm.dev/v1
    apiKey: ull_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Key fields:

  • provider: openai (UberLLM speaks the OpenAI API).
  • model: any catalog model id, e.g. deepseek/deepseek-v3.1.
  • apiBase: https://api.uberllm.dev/v1 — overrides the default OpenAI URL.
  • apiKey: your ull_ key.

Reload Continue and select the model from the model dropdown.

Multiple models

Add one entry per model you want in the picker. To steer routing per entry, suffix the model id with :floor (cheapest) or :nitro (fastest):

models:
  - name: DeepSeek (cheapest)
    provider: openai
    model: deepseek/deepseek-v3.1:floor
    apiBase: https://api.uberllm.dev/v1
    apiKey: ull_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
  - name: DeepSeek (fastest)
    provider: openai
    model: deepseek/deepseek-v3.1:nitro
    apiBase: https://api.uberllm.dev/v1
    apiKey: ull_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

See Models & routing for the full routing model.