First-class passport · hashes on the ledger, artifacts stay off-chain.
Aegis H100 Pool
OnlineOpen sourceAegis Mesh · Amara Voss
aegis/h100-pool
Spare GPUs into the pool. Earn the compute share of matched jobs.
Spare Aegis Mesh H100 capacity listed into the Cautus Intelligence compute pool. Listing compute is free; they earn the compute share of matched jobs.
GPU-only listing. Cautus Intelligence does not take a second platform cut when this capacity is matched to a model_only job — the call already paid the 6% rail take once.
Weights and inference stay on the host. Cautus Intelligence meters the gateway; buyers never receive the raw endpoint credentials.
H100 pool · matched to model_only listings
Endpoint registered (proxied): https://pool.aegis-mesh.internal/v1
Last heartbeat 10/11/2026, 4:19:10 PM
Hardware only — no model weights.
Lineage a buyer can click: base weights, license, fine-tune story, region, and custody. We do not invent eval scores. If no third-party card exists, the listing says so and still pins a card hash.
GPU-only listing. Cautus Intelligence does not take a second platform cut when this capacity is matched to a model_only job — the call already paid the 6% rail take once.
Mode: Open API
Compute listing. No platform take on the listing itself.
No approval queue. Sign in for credits, create an mg_ key, and call this slug. Competitor domain blocks still apply if the host configured them.
This call → usage tx → settlement tx on Litheum. Platform take on this listing is 0.0% (Compute only). Remainder goes to the model owner and, on matched jobs, the compute provider. Example at $0 / $0 per 1M: a 1M in + 1M out call costs $0.
Open Litheum explorer