Fix dependency conflict, remove deprecated models, and align credential validation #2517
Conversation
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
…`google-genai>=1.56.0` Fixes #2426
Remove 35 model configuration files that have been officially shut down or deprecated according to the Gemini API changelog (as of 2026-01-29): LLM models removed: - Gemini 1.0 series (gemini-pro, gemini-pro-vision) - Gemini 1.5 series (all variants including experimental) - Gemini 2.0/2.5 preview and experimental models - Legacy experimental models (gemini-exp-*) Text embedding models removed: - embedding-001 (shut down Oct 2025) - text-embedding-004 (shut down Jan 14, 2026) Also updated: - IMAGE_GENERATION_MODELS in llm.py to remove shut down models - text_embedding/_position.yaml to remove deleted model references Models retained for future deprecation monitoring: - gemini-2.0-flash series (shutdown: March 31, 2026) - gemini-2.5-pro/flash stable versions (shutdown: June 17, 2026) - gemini-3 preview series (no shutdown date announced) Closes #2515
Replace `count_tokens` API call with `generate_content` in validate_credentials method to align with other providers (OpenAI, Anthropic). This ensures credentials are validated against actual LLM usage, preventing cases where credentials pass validation but fail during real generation calls. Changes: - Use genai_client.models.generate_content() instead of count_tokens() - Add temperature=0 and max_output_tokens=20 config for minimal usage - Update docstring to clarify the validation approach Closes #2516
Contributor
|
Warning You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again! |
dosubot
bot
added
the
size:XXL
…g behavior Update test expectations to align with the implementation that filters out assistant messages with empty parts. Empty content assistant messages are now correctly expected to be removed rather than kept with empty parts lists.
Stream29
requested changes
Jan 29, 2026
Collaborator
Stream29
left a comment
Stream29
left a comment
There was a problem hiding this comment.
We should remove the models' names from _position.yaml without deleting the models file.
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.
Related Issues or Context
Fixes #2426
Closes #2515
Closes #2516
This PR contains Changes to Non-Plugin
This PR contains Changes to Non-LLM Models Plugin
This PR contains Changes to LLM Models Plugin
Version Control (Any Changes to the Plugin Will Require Bumping the Version)
VersionField, Not in Meta Section)Dify Plugin SDK Version
dify_plugin>=0.3.0,<0.6.0is in requirements.txt (SDK docs)Environment Verification (If Any Code Changes)
Local Deployment Environment
SaaS Environment