Edgework

Edgework is Forkjoin.ai's Cloudflare-facing app shell, API surface, browser transport layer, and proof-backed Aeon delivery surface — inference as a priced workload, with settlement receipts. Source: apps/edgework-app.

What this repository is

This is a model card, not a model. There are no weights here. Edgework fronts serving; the weights it can route are the admitted G1L knots published at forkjoin-ai/rwkv7-g1l-{1.5b,2.9b,7.2b,13.3b} (R2 is the serving copy).

Contract

Admission gate is routing authority. A knot is served only with a two-sided admission receipt; otherwise the request is refused.
Bounded surfaces. /api/knotgraft (Prismatic Refraction ingest) returns knotgraph-shaped blocks with tenant isolation enforced at emit time; per-tier rate limits are enforced in the worker.
Settlement. Inference is priced and receipted; there is no platform-paid AI path.

Model cards for the served knots carry the measured performance and admission evidence; this card exists so the serving surface is discoverable from the model organization.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support