fix(openrouter): support llama-index-llms-openai-like 0.7 - #21891
Merged
logan-markewich merged 2 commits intoJun 9, 2026
Merged
logan-markewich merged 2 commits into
logan-markewich merged 2 commits into
Conversation
The openrouter integration capped llama-index-llms-openai-like at <0.7, which transitively pinned llama-index-llms-openai<0.6 and blocked upgrading llama-index in environments that pull openrouter. Bump the openai-like cap to >=0.7.0,<0.8 (matching the current 0.7.x release line, which depends on llama-index-llms-openai>=0.7.0,<0.8) and the core lower bound to >=0.14.3 to match openai-like's floor. OpenRouter is a thin OpenAILike subclass; the OpenAILike constructor kwargs used here (api_base, api_key, model, additional_kwargs, etc.) are stable across the 0.5 -> 0.7 jump, so no code changes are needed. Patch the package version to 0.5.1. Fixes run-llama#21860
Companion to the previous commit: refresh the package's uv.lock so the locked transitive resolution matches the new dependency cap. The locked versions are llama-index-llms-openai-like==0.7.2 and llama-index-llms-openai==0.7.9, both already on PyPI.
gaurav0107
marked this pull request as ready for review
June 5, 2026 22:19
logan-markewich
approved these changes
Jun 9, 2026
Ricardo-M-L
pushed a commit
to Ricardo-M-L/llama_index
that referenced
this pull request
Aug 17, 2026
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
llama-index-llms-openrouter0.5.0 capsllama-index-llms-openai-like<0.7, which transitively pinsllama-index-llms-openai<0.6and blocks upgradingllama-indexin environments that pull the openrouter integration (#21860).This PR widens the cap to
>=0.7.0,<0.8(matching the current 0.7.x release line ofopenai-like, which itself requiresllama-index-llms-openai>=0.7.0,<0.8) and bumps thellama-index-corelower bound to>=0.14.3to matchopenai-like's floor. The package version is patched from0.5.0to0.5.1.OpenRouteris a thinOpenAILikesubclass that only sets a default API base and a fewadditional_kwargsfor provider routing. TheOpenAILike.__init__keyword arguments used here are stable across the 0.5 → 0.7 jump, so no code changes are required.Fixes #21860
Type of Change
How Has This Been Tested?
uv syncresolves cleanly with the new constraints inllama-index-integrations/llms/llama-index-llms-openrouter.uv run --group dev -- pytest tests/— both existing tests pass (test_embedding_class,test_provider_routing_settings_injected); the latter exercises the inheritedOpenAILikeconstructor end-to-end.uv run --group dev -- ruff check .— clean.Suggested Checklist:
Version Bump?
llama-index-llms-openrouter0.5.0→0.5.1(patch, dependency-only).New Package?