Key Takeaways
- n8n Cloud v2.36 introduces Gateway credits, letting supported AI model and tool nodes run without separate provider API keys.
- Six model providers and five tool services draw from one prepaid balance, tracked by workflow and by service inside the Cloud Admin Panel.
- Gateway credits follow provider list pricing with no markup, but self-hosted instances will not receive them until a later release.
Table of Contents
The API Key Bottleneck Just Disappeared for n8n Cloud Builders
The official n8n blog has broken the news that supported AI models and tool services no longer require separate provider credentials.
With the v2.36 release now rolling out to select n8n Cloud plans, teams can choose Gateway credits instead of adding their own API keys.
That means a shared prepaid balance sits inside n8n, ready to power six model providers and five tool services without breaking the workflow canvas.
Starting with v2.36, when you add a node for a supported AI model or service on select n8n Cloud plans, everything you need to run it is already in n8n.
As n8n’s official announcement details, the immediate impact is simple: experimentation no longer stalls on account signups, payment forms, or credential setup.
Inside Gateway Credits: Provider Coverage and Billing Controls
Gateway credits appear directly in the node credential selector as an alternative to bringing your own provider authentication.
Usage draws from one balance that is displayed in real currency, not abstract tokens or per-provider wallets.
The Cloud Admin Panel tracks consumption by workflow and by service, supports manual top-ups, and can enable automatic funding.
Instance owners can disable Gateway credits workspace-wide, giving teams a central control point for spend.
If a user describes a supported service to n8n Assistant and no credential exists, the assistant can apply Gateway credits automatically.
The node option now appears alongside fixed and end-user OAuth credential types in n8n’s credential documentation, and credential selection can be driven by expressions for reusable workflows.
When Gateway credits are used, the selected provider receives the content of the request—such as a prompt or document—but not the user’s identity or n8n account details.
Six model providers are available at launch.
- Anthropic — access to frontier Claude models.
- Google Gemini — Google’s multimodal model family.
- OpenAI — OpenAI models for general and agentic workloads.
- Alibaba Cloud / Qwen Cloud — open-source model family for coding and research agents.
- Kimi.ai — long-horizon coding and research with thinking and instant modes.
- MiniMax — M3 for coding and agentic tasks across a 1M-token context, H3 for video generation.
Five tool services round out the initial catalog.
- Brave Search — real-time web results from an independent 40-billion-page index.
- Firecrawl — turn websites into LLM-ready data while handling JavaScript, rate limits, and proxies.
- Browserbase — cloud browser agents that navigate, click, type, and extract structured data.
- LlamaParse — layout-aware OCR that preserves tables and headings in complex documents.
- PDF.co — generate, convert, extract, merge, split, and fill PDFs in a single node.
A competitive intelligence workflow can move from Brave Search to Browserbase for rendering interactive pages, then into LlamaParse for document structure, and finally to a supported model for synthesis.
That entire pipeline now happens without leaving the n8n editor or opening a separate vendor dashboard.
Prepaid Model Access Reshapes Automation Economics
Gateway credits attack a hidden cost in AI automation: the operational drag of managing accounts, invoices, and API keys across a growing roster of vendors.
For teams already running dozens of workflows, that sprawl quietly becomes its own maintenance job.
A single balance with per-service and per-workflow tracking shifts spend control back to the instance administrator.
Because n8n bills per workflow execution rather than per step, estimating the cost of a Gateway credit workflow remains relatively straightforward.
Existing AI credits tied to n8n’s in-app Assistant remain distinct from Gateway credits, which cover third-party model and tool nodes directly.
This separation eliminates a long-standing ambiguity in platform comparisons and procurement reviews.
n8n’s execution-based pricing also deserves attention: the Business tier has been positioned at €667 per month for 40,000 executions, with overage priced at €4,000 per additional 300,000-execution bucket.
Starter plans, by contrast, begin at €20 per month billed annually for 2,500 executions.
Those numbers matter because Gateway credits do not introduce a premium markup; n8n’s stated aim is to follow each provider’s published list price on a single public rate page.
For enterprise buyers, n8n Cloud’s EU hosting footprint on Azure—with AES256 and FIPS 140-2-compliant encryption—adds compliance weight to the single-balance model.
It also means finance teams can review one vendor instead of reconciling multiple provider invoices.
Self-hosted instances will not receive Gateway credits in this release, though n8n says that infrastructure work is on the roadmap.
The broader market signal is clear: workflow platforms are consolidating AI spend into a single prepaid balance, shifting the locus of vendor lock-in from individual model providers to the orchestration layer.
For n8n’s competitors, the pressure is now on to match this shortcut from idea to first successful run without forcing credential sprawl.
From Canvas to Production Without Leaving the Editor
Gateway credits collapse the distance between wanting to test a model and getting a successful first run, removing the setup tax that kills promising automation ideas. For teams building AI automation pipelines that need to scale, programmatic SEO and AI automation is how Andres SEO Expert approaches it — contact us to plan your next workflow.
Frequently Asked Questions
What are n8n Gateway credits and how do they work?
Gateway credits are a prepaid balance built directly into n8n Cloud that lets you run supported AI models and tool services without creating separate provider accounts or API keys. Starting with the v2.36 release on select n8n Cloud plans, Gateway credits appear in the node credential selector as an alternative to bringing your own authentication. Usage draws from one shared balance displayed in real currency, and if no credential exists when a supported service is requested, the n8n Assistant can apply Gateway credits automatically.
Which AI model providers and tool services are included with Gateway credits?
Gateway credits launch with six model providers: Anthropic (frontier Claude models), Google Gemini, OpenAI, Alibaba Cloud / Qwen Cloud, Kimi.ai, and MiniMax (M3 for coding and agentic tasks with a 1M-token context, plus H3 for video generation). Five tool services round out the catalog: Brave Search, Firecrawl, Browserbase, LlamaParse, and PDF.co. Together they cover search, web scraping, cloud browser automation, document parsing, and PDF generation inside a single n8n workflow.
Are Gateway credits more expensive than using your own API keys?
No premium markup is applied. n8n states its aim is to follow each provider published list price on a single public rate page, so Gateway credits are positioned as a convenience layer rather than a margin layer. Because n8n bills per workflow execution rather than per step, estimating the cost of a Gateway credit workflow stays relatively straightforward. Separate execution pricing applies: the Business tier has been positioned at 667 EUR per month for 40,000 executions with overage at 4,000 EUR per additional 300,000-execution bucket, while Starter plans begin at 20 EUR per month billed annually for 2,500 executions.
How do administrators track and control Gateway credit spending?
The Cloud Admin Panel tracks consumption by workflow and by service, supports manual top-ups, and can enable automatic funding so workflows do not stall. Instance owners can also disable Gateway credits workspace-wide, giving teams one central control point for AI spend. Because everything runs against a single balance instead of per-provider wallets, finance teams can review one vendor instead of reconciling multiple provider invoices.
Are Gateway credits available on self-hosted n8n instances?
Not in this release. Self-hosted instances will not receive Gateway credits with v2.36, though n8n says the required infrastructure work is on the roadmap. Gateway credits are currently limited to select n8n Cloud plans. On the Cloud side, the EU hosting footprint on Azure with AES256 and FIPS 140-2-compliant encryption adds compliance weight for enterprise buyers evaluating the single-balance model.
How do Gateway credits differ from n8n AI Assistant credits?
Existing AI credits tied to the in-app n8n Assistant remain distinct from Gateway credits. Assistant credits power the assistant experience itself, while Gateway credits cover third-party model and tool nodes directly inside workflows. This separation eliminates a long-standing ambiguity in platform comparisons and procurement reviews, since buyers can now see exactly which balance funds which capability.
What data is shared with model providers when using Gateway credits?
When Gateway credits are used, the selected provider receives the content of the request, such as a prompt or a document, but not your identity or your n8n account details. Requests are routed through the Gateway layer rather than a personal provider account, which keeps end-user credentials and account information out of the provider request path while still delivering the prompt payload the model needs to respond.
