Conversation
Agent-Logs-Url: https://github.com/rxtech-lab/markitdown-server/sessions/ab39b866-b695-4aae-8b60-d3420c894009 Co-authored-by: sirily11 <32106111+sirily11@users.noreply.github.com>
There was a problem hiding this comment.
Pull request overview
Updates the service’s default OpenRouter model and bumps the markitdown dependency version to a newer release.
Changes:
- Switch default LLM model from
google/gemini-2.5-flash-preview-05-20togoogle/gemini-3-flash-preview - Bump
markitdowninrequirements.txtto0.1.5 - Raise the
markitdown[all]lower bound inpyproject.tomlto>=0.1.5
Reviewed changes
Copilot reviewed 2 out of 3 changed files in this pull request and generated 2 comments.
| File | Description |
|---|---|
| converter.py | Updates the default model identifier used for conversions. |
| requirements.txt | Pins markitdown to 0.1.5 for the Docker/pip install path. |
| pyproject.toml | Updates the Poetry/PEP 621 dependency constraint for markitdown[all]. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| dependencies = [ | ||
| "fastapi (>=0.115.12,<0.116.0)", | ||
| "uvicorn (>=0.34.2,<0.35.0)", | ||
| "markitdown[all] (>=0.1.2,<0.2.0)", | ||
| "markitdown[all] (>=0.1.5,<0.2.0)", | ||
| "openai (>=1.82.1,<2.0.0)", |
There was a problem hiding this comment.
pyproject.toml now requires markitdown[all] >=0.1.5, but poetry.lock still pins markitdown at 0.1.4 (see poetry.lock around the markitdown package entry). This will make poetry install fail or continue using an out-of-date lock; regenerate and commit an updated lockfile so it resolves to a version that satisfies the new constraint.
| api_key=os.environ["OPENAI_API_KEY"], | ||
| ) | ||
| model = "google/gemini-2.5-flash-preview-05-20" | ||
| model = "google/gemini-3-flash-preview" |
There was a problem hiding this comment.
The model name is hard-coded, which requires code changes for future model bumps or environment-specific overrides. Consider reading the model from an environment variable (with this value as the default) so deployments can switch models without a new release.
| model = "google/gemini-3-flash-preview" | |
| model = os.environ.get("OPENAI_MODEL", "google/gemini-3-flash-preview") |
Upgrades the LLM model and markitdown dependency to their latest versions.
Changes
converter.py:google/gemini-2.5-flash-preview-05-20→google/gemini-3-flash-previewrequirements.txt:markitdown==0.1.2→markitdown==0.1.5pyproject.toml: lower bound>=0.1.2→>=0.1.5