Skip to content

fix: DeepSeek V4 proxy model recognition — substring match instead of exact set match for reasoning_content injection #9015

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Soulter merged 2 commits into from
Jun 27, 2026
+4 −2
Merged

fix: DeepSeek V4 proxy model recognition — substring match instead of exact set match for reasoning_content injection #9015

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d4eeb56
fix: DeepSeek V4 proxy model recognition — substring match for reason…
irmia2026 Jun 25, 2026
f1fa251
fix: remove deepseek-chat/reasoner exclusion per review feedback
irmia2026 Jun 26, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions astrbot/core/provider/sources/openai_source.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -933,11 +933,13 @@ def _finally_convert_payload(self, payloads: dict) -> None:
"""Finally convert the payload. Such as think part conversion, tool inject."""
model = payloads.get("model", "").lower()
is_gemini = "gemini" in model
deepseek_reasoning_models = {"deepseek-v4-pro", "deepseek-v4-flash"}
_deepseek_v4_markers = ("deepseek-v4-pro", "deepseek-v4-flash", "deepseek-v4")
is_deepseek_v4_reasoning = (
model in deepseek_reasoning_models
any(marker in model for marker in _deepseek_v4_markers)
or "api.deepseek.com" in self.client.base_url.host
)
# deepseek-chat and deepseek-reasoner now point to V4 models (per official website)

# MiMo 推理模型（MiMo-V2.5-Pro / MiMo-V2.5 / MiMo-V2-Pro / MiMo-V2-Omni / MiMo-V2-Flash）
# 要求 assistant 历史消息必须回传 reasoning_content，否则返回 400
mimo_reasoning_models = {
Expand Down
Loading