Menu
Mattering Automations connector

An OpenRouter MCP where one key reaches hundreds of models and every call reports its cost

Connect one API key and chat completions against models from every major provider become workflow steps, with fallback model lists that keep a pipeline running through outages, provider routing preferences, and an automatic USD cost figure on every response. Six nodes, one paid, five free reads.

6 ready-made nodes One credential to connect Works with Claude, ChatGPT, and Matter

What the OpenRouter MCP connector does

The OpenRouter connector for Mattering Automations is 6 nodes around one idea: routing. OpenRouter: Chat Completion sends one OpenAI-compatible request to any model in the catalog by id (anthropic/claude-sonnet-4, openai/gpt-5, google/gemini-2.5-flash), with a fallback model list that falls through in order when the first choice is down or over quota, and provider routing preferences that control which upstream providers serve it. Around it sit five free reads: the model catalog, per-model provider endpoints, the provider directory, the key's own limits and usage, and a per-generation cost audit.

Every node speaks MCP, the open protocol for connecting AI assistants to tools, so Claude, ChatGPT, and any MCP-capable assistant can use them, alongside Matter, the AI assistant built into the platform. An assistant can browse the live catalog with OpenRouter's per-token prices, compare the providers serving a model, run a completion on whichever model fits, and report exactly what it cost.

Because Mattering is a work platform rather than middleware, model work lands somewhere useful by default: a comparison sweep can write each model's answer, cost, and latency as rows in a Mattering database, and a content pipeline with a fallback model list can keep drafting into Mattering Productivity tasks even while its first-choice model has an outage.

Cost accounting is the second pillar, and it is automatic. Every chat response carries data.usage.cost, the USD amount charged for that call, so workflows track spend per request with no opt-in. The catalog nodes carry OpenRouter's own per-token prices for choosing models by budget, and OpenRouter: Get Generation audits any past generation after the fact: exact total cost, native and normalized token counts, the provider that actually served it, and latency. Only the chat node bills; the other five nodes are free to call.

One connection, three surfaces

  • Visual automations. Drag OpenRouter nodes onto the canvas and wire them to triggers, conditions, and other apps.
  • AI agents, yours or ours. These nodes speak MCP, the open protocol, so Claude, ChatGPT, and any MCP-capable assistant can use them, and Matter, the agent built into the platform, uses them natively.
  • The whole platform. Combine with native tasks, projects, and other connectors, no middleware in between.

Every OpenRouter node, ready to drop on the canvas

6 nodes, generated from the connector itself so this list always matches what ships. Expand a category to browse.

Read and query nodes 5

OpenRouter: Get Current API Key

Read the limits and usage of the API key making the call: data.limit and data.limit_remaining (USD credit limit, null when unlimited), usage totals for the current day, week, and month, the free-tier flag, and a masked display label. It never returns the full key material. Free to call and read-only, which makes it the connector's credential health check: a bad key gets a 401, a good key gets a 200 at no cost. The endpoint takes no parameters, so the config panel is intentionally empty.

OpenRouter: Get Generation

Fetch the full billing and performance record for one completed generation by id (the data.id returned by the chat node): data.data carries total_cost in USD, native and normalized token counts (prompt, completion, reasoning, cached), any cache_discount, the provider_name that actually served it, generation_time and latency in milliseconds, and the finish reason. Free to call; this is the cost-audit companion to every paid completion. Note: the record can lag the completion by a moment, so retry briefly on a 404 right after a call.

OpenRouter: Get Model Endpoints

List every provider endpoint serving one model: each entry at data.data.endpoints[] carries the provider_name, its own per-token pricing, context_length, max_completion_tokens, quantization, uptime, and recent latency/throughput percentiles. This is the data behind provider routing: use it to pick the order or only lists for the chat node's provider preferences. Free to call.

OpenRouter: List Models

List the OpenRouter model catalog: each entry at data.data[] carries the model id (author/slug form, ready for the chat node), name, description, context_length, per-token USD pricing (pricing.prompt and pricing.completion), supported parameters, and modality info. Free to call. Honest probe note: this endpoint answers 200 even without authentication, so it cannot verify your key; use the Get Current API Key node for that.

OpenRouter: List Providers

List every upstream provider OpenRouter can route to: each entry at data.data[] carries the provider's name, slug (the value used in the chat node's provider order / only / ignore lists), headquarters and datacenter country codes, and links to its privacy policy, terms, and status page. Free to call. The endpoint takes no parameters, so the config panel is intentionally empty.

Action nodes 1

OpenRouter: Chat Completion

Send a chat completion to any model on OpenRouter with one OpenAI-compatible request: pick the model by id (e.g. anthropic/claude-sonnet-4 or openai/gpt-5), optionally list fallback models and provider routing preferences, and read the answer at data.choices[0].message.content. Every response automatically includes data.usage.cost, the USD amount charged for the call, so downstream nodes can track spend per request. Billed per token at the routed model's price. Streaming is always off: the node returns one JSON response.

Connecting OpenRouter takes one credential

OpenRouter issues API keys from the Keys page in its settings (openrouter.ai/settings/keys). Store an ordinary inference key, not a management or provisioning key: those unlock admin surfaces this connector deliberately does not ship. Mattering sends the key as a bearer credential on every request. Billing is prepaid credits, spent per token at the routed model's price.

1

Create an API key on the Keys page in your OpenRouter settings. An ordinary inference key is the right kind; management and provisioning keys are for admin surfaces the connector does not use.

2

In Mattering Automations, open Connectors, click Add Connector, choose OpenRouter, and create a credential with that key.

3

Click Test Connection. Mattering reads the key's own record: a free, read-only call that returns the key's credit limit, remaining limit, and usage totals. The key label in the response is masked; full key material is never returned.

4

All 6 OpenRouter nodes appear in the canvas palette.

Common uses for the OpenRouter MCP

One chat node for every model

OpenRouter: Chat Completion takes one OpenAI-compatible request and runs it against any model in the catalog by id, so switching from one provider's model to another is a config change, not a new integration. The model field is free text on purpose: the catalog holds hundreds of models and changes weekly, and new ones work the day they ship.

Fallback lists that survive outages

Give OpenRouter: Chat Completion a fallback model list and the request falls through it in order when the first model is down or over quota. The response's data.model reports which id actually served, so a workflow can log when it ran on plan B. Provider routing preferences (order, only, ignore, price sorting) control which upstream providers handle the call.

Catalog and provider intelligence, all free

OpenRouter: List Models returns the live catalog with each model's context length and OpenRouter's per-token prices, OpenRouter: Get Model Endpoints breaks one model down into the provider offers serving it with pricing, uptime, and latency percentiles, and OpenRouter: List Providers supplies the provider slugs the routing preferences use. Three free reads that make routing choices informed instead of guessed.

Spend tracking that writes itself

Every chat response already carries data.usage.cost in USD, and OpenRouter: Get Generation turns any completion's id into a full billing record: exact total cost, prompt, completion, reasoning, and cached token counts, any cache discount, the serving provider, and latency. Route those numbers into a Mattering database and per-run cost reporting is a side effect, not a project.

Credential and budget health checks

OpenRouter: Get Current API Key reads the limits and usage of the key making the call: credit limit, remaining limit, and usage totals for the day, week, and month. It is free, read-only, and never returns full key material, which makes it both the connection test and a scheduled budget watchdog that can flag a workspace before credits run dry.

Example: A model comparison sweep that scores answers and costs into a Mattering database

Built on the Automations canvas in a few minutes, or described in plain language to Matter and built for you.

1

Trigger: a scheduled run carries a benchmark prompt and a shortlist of model ids drawn from OpenRouter: List Models, where OpenRouter's per-token prices already suggest which candidates fit the budget.

2

OpenRouter: Chat Completion runs the prompt against each shortlisted model id; each response returns the answer, its data.usage.cost in USD, and the data.model that served it.

3

OpenRouter: Get Generation takes each response's id and returns the exact total cost, token counts, serving provider, and latency for the call.

4

Each model's answer, cost, provider, and latency land as a row in a Mattering database, building a comparison table the team can sort by quality per dollar.

5

OpenRouter: Get Current API Key closes the run with the key's remaining limit, so the sweep reports what it spent and what is left.

6

A summary lands on a task in Mattering Productivity flagging the best answer and the cheapest acceptable one.

OpenRouter MCP questions, answered

What OpenRouter operations are supported?

6 nodes: chat completions against any catalog model with fallback model lists and provider routing preferences, the model catalog with OpenRouter's per-token prices, per-model provider endpoints with uptime and latency data, the provider directory, the key's own limits and usage, and a per-generation cost audit. Deliberately not shipped: the legacy text-completions endpoint (absent from OpenRouter's current API spec), media modalities like images and audio, and the management planes for keys, analytics, and org administration.

Does this work with Claude, ChatGPT, or other AI assistants?

Yes. Mattering exposes these nodes over MCP, an open protocol for connecting AI assistants to tools, so any MCP-capable assistant, including Claude and ChatGPT, can connect to your workspace and use them. Matter, the assistant built into the platform, uses the same nodes natively.

How do fallback models work?

OpenRouter: Chat Completion accepts a list of model ids alongside the primary model. OpenRouter tries them in order: if the first is down, over quota, or otherwise unavailable, the request falls through to the next. The response's data.model reports which id actually served, so workflows can log fallbacks rather than discover them in the bill. Provider routing preferences work alongside this, controlling which upstream providers serve the chosen model via order, only, and ignore lists and price sorting.

How does cost tracking work?

Automatically, at three layers. Every chat response includes data.usage.cost, the USD amount charged for that call, with no opt-in required. OpenRouter: Get Generation audits any past generation by id: exact total cost, native and normalized token counts, cache discounts, the serving provider, and latency. And OpenRouter: List Models carries OpenRouter's own per-token prices for every model, so workflows can pick by budget before spending anything. The connector never invents prices; every figure is OpenRouter's own data.

Why is there no credits or activity node?

Because they would always fail with the key this connector stores. OpenRouter gates its credits and activity endpoints behind a separate management key class, distinct from the ordinary inference key the credential holds, so a node for them would be a trap that returns errors with a perfectly good credential. The balance signal is not lost: OpenRouter: Get Current API Key returns the key's credit limit, remaining limit, and daily, weekly, and monthly usage totals on the ordinary inference key, for free.

Which models are available, and why is the model field free text?

Everything in OpenRouter's catalog: hundreds of models from every major provider, addressed by id in author/slug form. The field is free text rather than a fixed list because the catalog changes weekly, so new models work the day OpenRouter lists them, with no connector update. OpenRouter: List Models returns the live catalog with prices and context lengths, leaving the model unset uses your account's default, and openrouter/auto asks OpenRouter to choose.

Why does the models list succeed even when my key is wrong?

Because OpenRouter serves its catalog without authentication, so a successful call to OpenRouter: List Models proves nothing about your credential. The real credential check is OpenRouter: Get Current API Key, which requires a valid key, costs nothing, and returns the key's limits and usage on success. That is exactly what the connection test runs.

Is streaming supported?

No, by design. The connector pins streaming off so every execution of OpenRouter: Chat Completion returns one complete JSON response, which is what canvas workflows and downstream nodes consume; a token stream has no meaningful shape on an automation canvas. Requests also carry Mattering attribution headers, which affect only app attribution on OpenRouter's rankings, never routing, pricing, or billing.

6 nodes waiting on the canvas

Put OpenRouter to work inside your workspace

Connect once, then use every node in visual automations and AI agent workflows across the Mattering platform.

All product names, logos, and brands are property of their respective owners and are used for identification purposes only. Use of these names, logos, and brands does not imply endorsement or affiliation. OpenRouter is a product of its respective owner; Mattering is an independent platform.

Start Free Explore our Products