Reselling Venice (Diem) capacity

Venice subscription tiers (including Diem) include API access. If you have spare headroom, you can resell that capacity through Morpheus: your proxy-router forwards inference to Venice, and consumers pay you in MOR via marketplace bids. The simplest path today is MyProvider + an existing marketplace model Id — not minting a new model for every Diem holder.

Read Venice’s TOS. Resale terms vary by tier and over time — confirm yours allows reselling API access before going live. This page describes mechanics, not legal permission.

Why this path

Goal	Approach
Earn MOR from unused Venice quota	Bid on an existing Morpheus model that matches what Venice can serve
Avoid marketplace clutter	Do not register a new model if one already exists on active.mor.org
Stay graphical	Use MyProvider onboarding → Venice Diem resale path
Keep ops cheap	Docker/binary P-Node (or SecretVM if you want TEE for the router only — you still cannot tag Venice backends `tee`)

Architecture

Simple recipe (MyProvider-first)

Open MyProvider onboarding

Go to myprovider.mor.org. Choose Venice Diem resale in New provider onboarding.

Look up the model to join

Search active.mor.org (built into the wizard) for a model name Venice can serve. Copy the existing Id. Prefer bidding on that Id over creating a new model.

Fill Venice backend + secrets

Pick a Venice preset (chat / embeddings / TTS), paste your Venice API key, and export MODELS_CONFIG_CONTENT (and the rest of your .env / container secrets). Details also in MyProvider GUI.

Run a proxy-router

Docker or binary is enough — see Container P-Node. Point models-config / MODELS_CONFIG_CONTENT at Venice URLs with your key. Expose :3333 publicly; keep admin private (or HTTPS for hosted MyProvider).

Connect MyProvider to your node → Provider tab → register host:3333. Models & Bids → Available Models → Add Bid on the Id from step 2. Do not use the tee tag.

Price above your Venice cost

See Pricing a resale bid. Every bid costs a non-refundable 0.3 MOR fee — finish config before posting.

Verify

Follow Verify your provider setup. Watch Venice usage limits so a busy session cannot blow your Diem quota.

Manual models-config (reference)

If you prefer editing JSON by hand:

{
  "$schema": "./internal/config/models-config-schema.json",
  "models": [
    {
      "modelId": "0x<existing_active_mor_org_model_Id>",
      "modelName": "matching-on-chain-name",
      "apiType": "openai",
      "apiUrl": "https://api.venice.ai/api/v1/chat/completions",
      "apiKey": "<your_venice_api_key>",
      "concurrentSlots": 4,
      "capacityPolicy": "simple"
    }
  ]
}

Restart the proxy-router after edits. Full register/bid contract steps: Register on chain.

Operational tips

Track Venice usage — set Venice account limits/alerts so a session can’t blow your budget.
Throttle concurrentSlots — start conservatively. Going too high causes upstream 429s.
Failover: multiple upstream accounts → multiple proxy-routers and separate bids.
Avoid the tee tag — you cannot prove anything about Venice’s runtime. Only use tee when you control the backend on a SecretVM-style TEE.

Get started

Inference API (hosted)

Concepts

Consumers

Prosumers

Providers - Resale

Providers - Full P-Node

AI knowledge

Ecosystem

Reference

Why this path

Architecture

Simple recipe (MyProvider-first)

Manual models-config (reference)

Operational tips

​Why this path

​Architecture

​Simple recipe (MyProvider-first)

​Manual models-config (reference)

​Operational tips

​Related

Why this path

Architecture

Simple recipe (MyProvider-first)

Manual models-config (reference)

Operational tips

Related