Skip to content

Update history compression summarize model to GPT-5.4 nano#3688

Merged
DavidBakerEffendi merged 1 commit into
masterfrom
3687-update-history-compression-summarize-model-from-retired-gemini-31-flash-lite-preview
May 27, 2026
Merged

Update history compression summarize model to GPT-5.4 nano#3688
DavidBakerEffendi merged 1 commit into
masterfrom
3687-update-history-compression-summarize-model-from-retired-gemini-31-flash-lite-preview

Conversation

@DavidBakerEffendi

Copy link
Copy Markdown
Collaborator

Summary

  • Switch the summarize/history compression default off the retired Gemini Flash Lite preview model and onto gpt-5.4-nano.
  • Add a regression test to lock ModelType.SUMMARIZE to the new default and prevent preview-model regressions.

Key Changes:

  • Updated ModelProperties so SUMMARIZE now resolves to the existing gpt5_4NanoLow config.
  • Added ModelPropertiesTest to assert the summarize default is gpt-5.4-nano and does not use a -preview model ID.

Touch Points:

  • app/src/main/java/ai/brokk/project/ModelProperties.java
  • app/src/test/java/ai/brokk/project/ModelPropertiesTest.java

Testing

  • ./gradlew fix tidy passed.
  • ./gradlew :app:test --tests ai.brokk.project.ModelPropertiesTest passed.
  • ./gradlew analyze passed.

@DavidBakerEffendi
DavidBakerEffendi marked this pull request as ready for review May 27, 2026 16:03
@DavidBakerEffendi DavidBakerEffendi self-assigned this May 27, 2026
@DavidBakerEffendi
DavidBakerEffendi merged commit acb3481 into master May 27, 2026
5 checks passed
@DavidBakerEffendi
DavidBakerEffendi deleted the 3687-update-history-compression-summarize-model-from-retired-gemini-31-flash-lite-preview branch May 27, 2026 16:29
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Update history compression summarize model from retired gemini-3.1-flash-lite-preview

1 participant