Plugin Type

VSCode Extension

App Version

4.119.4

Description

I've been testing indexing with a local ollama model, and it keeps failing with a connection error after exactly 60 seconds according to the logs. The indexing itself appears to work at the start most of the time but can fail after 5 or 10 minutes, only if a request runs up against the limit.

It looks like the ollama embeddings is hardcoded at 60 seconds in src/services/code-index/embedders/ollama.ts.

The fix would be to use the getApiRequestTimeout helper function so it respects the user's setting.

Reproduction steps

Run the Codebase Indexing on a large repo with a decent sized model that can be slow to respond.
Wait for the indexing to fail.

Provider

Ollama

Model

qwen3-embedding:8b-q8_0

System Information

No response