DeepSeek Harness + OpenBulu configuration
In Custom model API settings, enter the provider, base URL, protocol, credential, and exact model ID.
Protocol: OpenAI-compatible. Base URL: https://openbulu.com/v1. Endpoint: POST /v1/chat/completions.
Configuration reference
Settings โ Models โ Add model provider โ Custom model API
Provider ID: openbulu
Base URL: https://openbulu.com/v1
API protocol: OpenAI Chat Completions
Credential: <OPENBULU_API_KEY>
Model: <MODEL_ID>Configuration and verification steps
- In DeepSeek Harness, open Settings > Models, add a provider, and choose Custom model API. Use a lowercase provider ID such as openbulu.
- Enter https://openbulu.com/v1 as the base URL, choose OpenAI Chat Completions, add your OpenBulu key, and select an exact available model ID. Fetch available models if the gateway supports discovery, or add the ID manually.
- Save the provider, select its model in the composer, send a small request, and compare the time, model, and /v1/chat/completions route with OpenBulu usage logs. Test file and tool actions separately.
Use your own API key and an available model ID. Confirm the selected provider and model in the client, run a small request, and match its time, route, and model in your OpenBulu usage log. Advanced tool actions may need separate testing.