Repository navigation
docs: document LLMTR as a model endpoint - #178
Open
knowhycodata wants to merge 1 commit into
Open
knowhycodata wants to merge 1 commit into
knowhycodata wants to merge 1 commit into
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #177
Hi! I help maintain LLMTR, an OpenAI-compatible model gateway. Dot chat and research already talk to whatever endpoint is in
OPENAI_BASE_URL, so OpenDots works with LLMTR without any code changes. This PR just writes the setup down so people don't have to work it out themselves.What changed
.env.example: a commented-out LLMTR preset placed below the defaults. Node keeps the last assignment it sees, so uncommenting the preset overrides the defaults. The comment also says to leave only oneOPENAI_MODELline active.docs/SETUP.md: a short LLMTR section covering the key format (llmtr-...), how to choose a model id (a leadingllmtr/is part of the id and is passed through as is), and a one-line curl check.README.md: one sentence pointing to that section.Voice is unchanged and still uses the OpenAI Realtime API.
Testing
I started from a fresh clone of this branch on Node 24.22.1:
npm ci,npm run check-format,npm run lint,npm run typecheck,npm test(55 files, 346 tests) andnpm run buildall pass.npm run knipreports only findings that are already onmain.cp .env.example .env, uncommented the LLMTR lines withOPENAI_MODEL=qwen/qwen3.8-27b-freeand set a real key.npm startcame up, and/api/workspaceshowedmodel: true. The only missing item was the Intelligence key, which I don't have locally.DotAgent, asking it to create a page titled "LLMTR check" containing the word "pong". The model calledcreate_space_page, the page was created, and the reply streamed back. The turn took about 4 seconds and raised noRUN_ERROR.Happy to adjust the wording or move things around if you'd like it organized differently. Thanks for taking a look!