Use BlockRun as the OpenAI-compatible model behind n8n's AI nodes
Point n8n's OpenAI credential base URL at BlockRun's API-key endpoint and every workflow can call any catalog model from one credential, billed at exact usage on one invoice.
How it works
n8n runs in the cloud or on your own server, and its AI nodes take an OpenAI-compatible credential with a base URL. A hosted workflow cannot reach a sidecar on your laptop, so the door here is the invoiced one: an API key issued to your account, a base URL on BlockRun's API host, and every chat model on the catalog available to the OpenAI Chat Model node by id.
Usage is metered at exactly what each call used and billed in dollars monthly, with no per-call minimum — the shape a workflow that runs on a schedule needs. Self-hosted n8n can use the pay-per-call door through a sidecar on the same host instead; the node configuration is the same, only the base URL and key change.
Behind it is the BlockRun catalog: 78 chat models plus image, video, music and speech generation, search, market data and multi-chain RPC, priced per call. 6 open-weight models are free and need no wallet.
Setup
1. Create an OpenAI credential in n8n
Base URL: https://api.blockrun.ai/v1
API Key: <your BlockRun key>2. Use it in the OpenAI Chat Model node
Model: anthropic/claude-sonnet-5 (any catalog id)Questions
- How do I connect n8n to BlockRun?
- Create an OpenAI credential whose base URL is BlockRun's API host and whose key is the one issued to your account, then pick any catalog model id in the OpenAI Chat Model node.
- Does n8n Cloud work, or only self-hosted?
- Both. n8n Cloud uses the API-key door, since it cannot reach a local sidecar; self-hosted n8n can use either door.
- How do I get a BlockRun API key for n8n?
- Keys are issued to your account — see the teams page. Metering starts with the first call and the invoice is the ledger, totalled.
- Which models can an n8n workflow use?
- Every chat model on the catalog, by id, from one credential. Switching the model in a node is changing a string.
- Can n8n generate images or search the web through BlockRun?
- Yes, through the HTTP Request node against the image and search endpoints with the same key; the docs carry each request shape.
- n8n on BlockRun vs n8n with a provider key per vendor?
- One credential and one invoice for every vendor's models, instead of a credential and a bill per vendor, and a model switch that never needs a new credential.
- What does a call cost through this integration?
- The provider's published token rate with no markup on tokens, plus a flat per-call fee, quoted in dollars in the 402 before the call runs. Image, video and speech calls are priced per unit the same way. The live rate card is on the models page.