The best LLM for n8n
n8n turns AI into automation — an AI Agent node that reads data, calls tools, and acts, all without code. But an automation that stalls because a provider rate-limited mid-run is worse than no automation: it fails silently in the background. CodeBurst plugs into n8n's OpenAI Chat Model node through a custom Base URL and adds automatic failover and tool-call repair, so your workflows keep running.
Workflows fail quietly — so reliability matters more
An n8n AI Agent can loop, call tools, and chain steps, racking up model calls per execution. When one trips a provider's rate limit or returns an empty tool turn, the run errors out in the background — often unnoticed until a downstream step is missing data. A backend that routes around provider failures is exactly what an unattended automation needs.
What CodeBurst adds
| Failure | CodeBurst |
|---|---|
| Provider rate-limits mid-run | Reroutes to a healthy provider in the same request; the node returns normally. |
| Empty tool-synthesis turn | codeburst-agent retries with a corrected format. |
| A step needs more reliability | Set that node's model to codeburst-swarm for a multi-model vote. |
Set it up (no code)
- In n8n, go to Credentials → New → OpenAI.
- Paste your CodeBurst API key.
- Set Base URL to
https://codeburst.ai/api/v1(include the/v1). - Add an OpenAI Chat Model node using that credential; set the model to
codeburst-agent. - Connect it to your AI Agent (or Basic LLM Chain) node — done.
Because the OpenAI Chat Model node feeds every AI node, this single credential gives all of them provider failover and tool-call repair.
Get started
Get an API key How agent failover works →FAQ
How do I connect CodeBurst to n8n?
OpenAI credential → key + Base URL https://codeburst.ai/api/v1 → OpenAI Chat Model node (model codeburst-agent) → AI Agent.
Why use it in workflows?
Unattended runs stall silently on rate limits; CodeBurst fails over and repairs tool turns.
Works with the AI Agent node?
Yes — the OpenAI Chat Model node feeds all AI nodes.