🪲 BUG-#16: Fill in missing context-window entries - #35
Merged
Conversation
FernandoCelmer
commented
Aug 15, 2026
FernandoCelmer
left a comment
FernandoCelmer
left a comment
Member
Author
There was a problem hiding this comment.
The coverage expansion and case-insensitive matching are correct and well-tested. The test that asserts every templates/ default model resolves to a non-default entry is a strong safety net. One concern about the DEFAULT_CONTEXT_WINDOW reduction found below.
DEFAULT_CONTEXT_WINDOW had been lowered to 32_000 alongside adding the expanded model table, silently making any deployment on an unlisted model (a private/future model, a vendor-prefixed variant, a fine-tune) compact 4x more often than before with no warning. Restored the previous 128_000 default — still conservative relative to the smallest window in the table, just not a regression for existing unlisted-model setups.
Merged
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Fills in missing context-window entries and fixes case-insensitive model matching (issue #16). Adds every model family shipped in
templates/(Gemini, Grok, Groq, Kimi, DeepSeek, Qwen, NVIDIA/Together Llama, o3/o4). Matching is now case-insensitive to handle vendors with inconsistent casing (e.g. Together AI'sLlama-3.3-70B-Instruct-Turbo). LowersDEFAULT_CONTEXT_WINDOWto32_000so unrecognized future models compact early rather than risking provider rejection.Motivation and Context
Closes #16.
_CONTEXT_WINDOWSonly covered 10 model families; all others silently fell back to128_000— causing Gemini (1M real window) to compact far too early and smaller models to risk late compaction and provider rejection. Case-sensitive matching also missed mixed-case vendor model IDs.Types of changes
Checklist