feat: make qwen3.6-plus-free default model, add compaction models cascade#234
Merged
Conversation
Adding .gitkeep for PR creation (default mode). This file will be removed when the task is complete. Issue: #232
…ascade (#232) - Change DEFAULT_MODEL from minimax-m2.5-free to qwen3.6-plus-free - Add --compaction-models CLI option accepting links notation sequence - Implement compaction model cascade: tries models from smallest to largest context, skipping models whose context is too small or that hit rate limits - Default cascade: (big-pickle nemotron-3-super-free minimax-m2.5-free gpt-5-nano qwen3.6-plus-free same) - Update provider priority lists to include new free models - Add CompactionModelEntry interface for cascade entries Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add qwen3.6-plus-free and nemotron-3-super-free to free models docs - Update default model references from minimax-m2.5-free to qwen3.6-plus-free - Add context window sizes to FREE_MODELS.md table - Update MODELS.md pricing table and recommendation list Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add test for DEFAULT_MODEL being qwen3.6-plus-free - Add test for DEFAULT_COMPACTION_MODELS links notation sequence - Add test for getCompactionModelsFromProcessArgv - Add tests for CompactionModelConfig with cascade support - Verify backward compatibility without compactionModels array Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Document the motivation, solution design, context window data, and cascade logic for the free models update and compaction cascade. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
konard
changed the title
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Contributor
Author
🤖 Solution Draft LogThis log file contains the complete execution trace of the AI solution draft process. 💰 Cost estimation:
📊 Context and tokens usage:Claude Opus 4.6:
Total: 142.0K + 13.4M cached input tokens, 36.5K output tokens, $8.495370 cost Claude Haiku 4.5:
Total: 173.0K + 1.1M cached input tokens, 10.1K output tokens, $0.366347 cost 🤖 Models used:
📎 Log file uploaded as Gist (3152KB)Now working session is ended, feel free to review and add any feedback on the solution draft. |
Contributor
Author
✅ Ready to mergeThis pull request is now ready to be merged:
Monitored by hive-mind with --auto-restart-until-mergeable flag |
This reverts commit a43e46c.
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
1 participant
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
Fixes #232
minimax-m2.5-freetoqwen3.6-plus-free(~1M context window, 5x larger)--compaction-modelsCLI option accepting a links notation references sequence for cascading compaction models(big-pickle nemotron-3-super-free minimax-m2.5-free gpt-5-nano qwen3.6-plus-free same)— ordered from smallest/cheapest to largest contextqwen3.6-plus-free,nemotron-3-super-free) to provider priority lists and documentationdocs/case-studies/issue-232/How the cascade works
During compaction, the system tries each model in order:
sameas the final entry falls back to the base modelFiles changed
defaults.ts,argv.ts,run-options.js,model-config.jscompaction.ts,prompt.ts,message-v2.tsprovider.ts(priority lists)FREE_MODELS.md,MODELS.md,README.md,docs/case-studies/issue-232/compaction-model.test.ts.changeset/update-free-models-232.mdTest plan
--compaction-model(singular) still worksCompactionModelConfigwithoutcompactionModelsfield works (backward compat)