Summary
According to a post on X by @QwenAIFans, ggml-org tagged llama.cpp and ggml as v1.0.0 on Nov 13, ending 3.5 years of build-numbered releases. Breaking C API changes will now come only with major-version bumps. The release removes 35 deprecated functions and adds FP8 KV caches, a host-memory MoE expert cache, and security hardening for llama-server.
AI-assisted summary based on the listed source.
What happened
After 3.5 years of build-numbered releases, ggml-org tagged llama.cpp and ggml v1.0.0 on Nov 13; breaking C API changes will now come only with major-version bumps. The release removes 35 deprecated functions and adds FP8 KV caches, a host-memory MoE expert cache and llama-server security hardening. Why it...
Why it matters
Developers building local and private AI apps on llama.cpp can pin to a more stable API, since breaking C API changes are now tied to major versions. Teams that rely on the 35 removed deprecated functions will need to update their code before upgrading. Watch for the full release notes to confirm the details in the post.
Signal Intelligence
Signal Strength 95%
Technical label WATCH
Public Interest 71
Category AI AT WORK
Reader Depth TECHNICAL
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 82
Practical Impact Score 82
Novelty Interest Score 94
Consequence Score 50
Curiosity Score 0
Shareability Score 78
Why this is here
VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Scobleizer News Pulse (X).