UberLLM

Cline

Cline (the VS Code agent) has a built-in OpenAI Compatible provider that works with UberLLM directly.

Setup

  1. Click the ⚙️ settings icon in the Cline panel.
  2. Set API Provider to OpenAI Compatible.
  3. Fill in the fields:
    • Base URL: https://api.uberllm.dev/v1
    • API Key: your UberLLM key (ull_...)
    • Model: a catalog model id, e.g. deepseek/deepseek-v3.1
  4. Save. Cline will use UberLLM for all completions.

Model configuration (optional)

Under Model Configuration you can set context window, max output tokens, and image support. Match these to the model you selected — the UberLLM catalog lists each model's context_length and max_output_tokens at GET https://api.uberllm.dev/v1/models and on the Models page. Leaving Cline's defaults is fine for most models.

Tips

  • Pick a model with tool support for agentic workflows — Cline relies on function calling. The catalog marks supports_tools.
  • To always take the cheapest healthy provider for a model, append :floor to the model id (e.g. deepseek/deepseek-v3.1:floor). See Models & routing.
  • Watch spend in Dashboard → Activity on uberllm.dev; set a per-key spend limit when you create the key if you want a hard ceiling.