Back to AI intel
重点

llama.cpp b9590 Fixes LFM2/LFM2.5 json_schema Handling

AI intel briefing

Core summary

One sentence to understand this update

llama.cpp release b9590 addresses a bug where LFM2/LFM2.5 specialized template handlers incorrectly ignored `json_schema` from the response format.

Impact & opportunity

What this could mean

Developers relying on `json_schema` with LFM2/LFM2.5 in llama.cpp can now expect correct behavior and reliable tool-calling.