Go from workflow idea to first run without stopping to create provider accounts, generate API keys, or set up separate billing.
It’s happened to all of us. You find a template you wanted to try, open the first node, and then had to leave n8n to create an account with a provider, find the API keys page, add billing, copy the key, and come back to set up a credential.
Borrowing a favorite term from my grandmother, now you can just “skip it”.
The team is rolling out a way to use supported AI models and services directly on n8n Cloud using Gateway credits. On supported nodes, instead of adding a credential, just choose Gateway credits instead and be on your way. You don’t need a separate provider account or API key, usage comes out of a balance for the instance that you can top up.
Availability: For the initial release, n8n Cloud self-serve plans, including trials, on n8n 2.36 or later. We’re rolling out in phases, so if you don’t see it in your n8n Cloud accounts today, its on the way.
What you can reach for today
- AI models: OpenAI, Anthropic, Google Gemini, Alibaba Cloud Qwen, MiniMax, Moonshot Kimi
- Tool services: Brave Search for live web, news and local results from its own independent index, Firecrawl for turning a page or a whole site into clean structured content, Browserbase for when something has to click through a site rather than just read it, LlamaParse for making sense of complex documents with tables and awkward layouts, and PDF.co for routine PDF conversion and extraction.
These eleven providers and services are just the start. We’re working on another curated batch, picking from the APIs and capabilities people use most.
Why we built this and how it works
Credential setup is one of the first walls people hit when getting started or trying a new service in n8n. It can also make experimenting feel like more of a commitment than it should. Sometimes you just want to see whether a different model works better for a task or test a few services in a workflow before making a decision.
The team has worked on making credential setup easier before with one-click OAuth on Cloud and simplified connections to 70+ MCP servers. This takes that a step further. When you choose Gateway credits, n8n uses a managed credential to send the request through its gateway to the provider. The provider receives the content of the request, such as the prompt you send to a model or the text you send to a parser, but not your identity or n8n account details.
For experimenting, that means you can try a model or service while you’re already building, without first deciding whether it’s worth creating another account, setting up billing, and managing another API key. If it turns out to be the right tool for the job, you can keep using credits and top up when you need to, or switch that node to your own credential later.
How to use it
-
In the Node panel, you’ll see a green flag next to supported models and tool services indicating you can use credits.
-
Once you add the node to your workflow, choose Gateway credits in the Credential setup.
That’s it. The choice is per node, so you can use Gateway credits for some services and your own credentials for others in the same workflow. You can also keep bringing your own credentials exactly as you do today. Gateway credits aren’t intended to replace that option. They’re another way to get started or use supported services without managing an account with each provider.
On pricing
Using n8n credits is prepaid: requests to supported models and services are deducted from your instance’s credit balance as you use them. When the balance gets low, the instance owner can add more funds so you can keep using credits.
We follow publicly listed provider prices wherever we can, and every rate lives on one public page: app.n8n.cloud/service-pricing. For services that don’t map as neatly to token-based pricing, such as Firecrawl and PDF.co, we’ve made our best effort to align with typical vendor pricing.
Eligible new Cloud accounts will start with a small credit balance. Existing paying Cloud customers, excluding Enterprise, will also receive a one-time allowance as the rollout ramps up. It’s a good chance to try a model or tool service you’ve been curious about without setting anything else up first.
Managing what you spend
Credits come from one balance shared across your n8n instance. In Manage > Gateway credits, you can see your remaining balance and break down your spend by model or workflow.
You can also set up automatic top-ups. Choose the balance that should trigger a top-up, how much you want to top back up to, and an optional monthly limit.
Instance owners can turn Gateway credits on or off for the workspace from Manage > Workspace > Workspace settings > Enable Gateway credits in the Admin dashboard.
Where this goes next
For now this is on Cloud. The bigger vision includes a path for self-hosted users, but connecting the services infrastructure to instances running outside n8n Cloud is a more involved challenge, so it isn’t part of this initial release.
We’re already starting on our next curated batch of models and services to enable with credits. Beyond that, we’re thinking about how we can scale this option to many more providers and give you more control over how credits are managed in an instance.
We’d love to hear what you’d like to see next, whether that’s a particular model or service, or how to improve your experience once you’ve tried it out.