OmenWorks/bend-heads

Bend heads — the decision-interface shelf

One model per size. Heads swap on the same loaded copy. A head is not a model: it is the small interface that turns a loaded base model into a mode (structured decisions of a given family). Pick a size model (bend-4b, …), pick a head, switch modes without reloading anything.

The heads

• kev-native — ships with bend-4b; the base's own decision mode. Chat + decisions from one load. Status: native, receipted (SYS1-DUAL: 28/28 byte-identical goldens; one process, 6,491 vs 12,597 MiB vs the two-copy baseline).
• clef-flash — Cloudflare's Clef family. Runs end-to-end on the Bend engine (parity |Δp| 0.0092 vs reference after one wire fix). Status: wired; GPU path int8 ≤0.013, top-1 24/24; MLX path runs.
• jev-r2 — TypeSafe's Jev family. 313/313 byte-precise runs (max |dp| 2.2e-6). Status: wired.
• clef-omni — Cloudflare's omni release (2026-10-09). Status: not yet run here — in queue.

The rule

Anything in the wired shape (a trained interface over a supported base — Qwen-family bases included) loads and runs with no per-head work. A genuinely new head architecture needs a small port (done once before, canonport-style — days, not a rewrite). The parity receipts above are the gate: a head is "listed" when its runs are checkable.

See SUPPORT-MAP.md for the regenerated, source-labeled map.