Qwen3.8-Flash-Next
Guild’s primary development target. Hybrid recurrent and attention execution with routed experts, per-layer embeddings, and an optional MTP draft path.
View compatibility & sources Model catalog / Core 140f6f7
Compatibility,
with context.
Two built-in registry models. Clear boundaries around what runs, what is planned, and what is still a research question.
Supported means Guild’s primary development target, not a guarantee for every quantization or
hardware configuration. Experimental entries have a development execution path. Planned
entries describe architecture work. Research entries have no released weights.
6 entries
Ornith 1.5 35B
A registered and benchmarked hybrid MoE model. Guild’s README classifies the Qwen3.5 MoE / Ornith execution path as experimental and under active development.
View compatibility & sourcesGLM-style MoE
Metadata inspection and planning descriptors exist. The pinned README lists execution support as planned; an archetype descriptor is not a complete inference runner.
View compatibility & sourcesDeepSeek-style MoE
Metadata inspection and planning descriptors exist. The pinned README lists execution support as planned; an archetype descriptor is not a complete inference runner.
View compatibility & sourcesMixtral-style MoE
Metadata inspection and planning descriptors exist. The pinned README lists execution support as planned; an archetype descriptor is not a complete inference runner.
View compatibility & sourcesEon
Guild Labs’ proposed model family. Research interests include efficient language models, sparse MoE architectures, and ternary representations.
View compatibility & sourcesNo matching entries. Try a different search or status.
Registry fields are extracted at build time from registry.cpp. Status follows the pinned README. Hub metadata checked 22 October 2026.