Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
C
release: 1.11.0 also carries #1457 and #1456
JustVugg3a1653f113c6
diff --git a/CHANGELOG.md b/CHANGELOG.mdindex 249343e9..0e1754b3 100644--- a/CHANGELOG.md+++ b/CHANGELOG.md@@ -5,7 +5,7 @@ Format follows [Keep a Changelog](https://keepachangelog.com/).## [1.11.0] — 2026-09-13-54 pull requests since v1.10.2. A ninth model family, five real bugs closed+56 pull requests since v1.10.2. A ninth model family, five real bugs closedacross four engines, and the two platforms the C tests never built on nowbuilding them in CI.@@ -99,6 +99,11 @@ building them in CI.- **#1447** (@trigger2k20): `coli mirror verify` accepts a complete mirrorcopied by other means, with no receipt, and fails closed on a short orcorrupt shard.+- **#1456** (@trigger2k20): the planner reports `memory.unified` from the+ host, not from whether the selected engine has a placement-capable GPU, so+ a CPU-only engine on Apple Silicon no longer reads as non-unified; and+ `coli tune` stops suggesting `DRAFT`/`PIPE`/`PIN`/`NUMA` to glm53, which+ ignores them.- **#1396** (@dmoraesrs): the web reasoning selector drops "Medium" forGLM 5.3, which the engine renders identically to "High".- **#1428**, **#1435** (@iiEliJas), **#1433** and **#1432** (@texasich),@@ -117,6 +122,9 @@ building them in CI.on an M4 Max: 0.469 to 0.621 tok/s at 32 tokens, up to 2.15 tok/s withthe expert cache tuned. Also fixes `coli run --cap N` being silentlyignored on the glm53 one-shot path.+- **#1457** (@trigger2k20): GLM-5.3-Flash feeds the shared routing telemetry+ and writes `.coli_usage`, so usage-driven placement and partial-mirror+ planning have data for it.- **#1454** (@yuripourre): the qwen36 GPU tier compiles with `HIP=1`.- **#1399** (@SebaWag): `TRUNK_RESIDENT_LAYERS=N` streams the dense trunkthrough mmap (#826).