scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-11-11 08:38:24 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 359

test(e2e): repin the Together reasoning replay backend to MiniMax-M3.1-Flash (#50601)

Browse source 
MiniMax-M3 reached its 2026-11-06 deprecation date in the cost map, which names MiniMax-M3.1-Flash as its successor
This commit is contained in:
yuneng-jiang 2026-11-09 11:03:39 -08:00 • committed by GitHub
parent 001126cd1c
commit 1d06a1ca69
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 3 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
tests/e2e/llm_translation/test_together_ai_e2e.py
View file

@ -7,9 +7,9 @@ backends are pinned because the registry has no flag for what they prove: ``enab
and the ``{"reasoning": {"enabled": false}}`` toggle that ``reasoning_effort="none"`` maps
to are Qwen hybrid-model contracts, and the structured-output case runs on that hybrid
model with reasoning off, since a reasoning-only model can spend the whole token budget
thinking and return no content. MiniMax-M3 is the serverless model whose
thinking and return no content. MiniMax-M3.1-Flash is the serverless model whose
template renders a replayed ``reasoning_content`` back into the prompt (Qwen and
DeepSeek silently drop it). MiniMax-M3 honors that replayed field on nearly every call, not
DeepSeek silently drop it). MiniMax-M3.1-Flash honors that replayed field on nearly every call, not
every call (one miss in dozens of otherwise identical calls), so the replay case asks
up to ``REPLAY_ATTEMPTS`` times and fails only when no answer carries the secret, which
a proxy that strips the field guarantees. Requires TOGETHER_API_KEY on the proxy; no
@ -56,7 +56,7 @@ from pydantic import BaseModel
pytestmark = pytest.mark.e2e
HYBRID_REASONING_BACKEND = "together_ai/Qwen/Qwen3.5-9B"
REASONING_REPLAY_BACKEND = "together_ai/MiniMaxAI/MiniMax-M3"
REASONING_REPLAY_BACKEND = "together_ai/MiniMaxAI/MiniMax-M3.1-Flash"
SECRET_PROMPT = "Remember this for later and reply with just OK."
SECRET_REASONING = "The user told me their favorite color is chartreuse. I must remember it."