Skip to content
Features
-
Persistent shared‑prefix KV cache for reusable prompt prefixes across sessions
-
Palimpsest: on‑disk n‑gram archive that drafts from previously generated content at zero cost
-
Il Suggeritore: grammar‑as‑draft source proposing constrained tokens during decoding
-
Self‑regulating speculative decoding that automatically disables when not beneficial
-
The Conclave: consensus‑based multi‑candidate generation to improve answer quality
Upgrade now
v0.1.6
Mixed
·
Breaking upgrade
Warm‑ahead fix + `/v1/warm`
No immediate action
v0.1.5
Mixed
·
Decoding downgrade + Qwen3.5-9B addition
No immediate action
v0.1.4
Bug fix
·
Chat endpoint respects templates
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
About
Languages
C++
·
CMake
·
Shell
View on GitHub
Search tools, categories, lists, and users
Use ↑↓ to navigate, Enter to open, Esc to close
No results for ""
⌘K to open
↑↓ navigate
⏎ open