easybits
Configure a task on the easybits platform, call it from n8n with our verified community node. No prompt chain to build, no parser to maintain.
One node instead of a chain. You configure the task on the platform, no code, nothing to wire in n8n.
Three recommended models per task. We research which ones work, so you do not have to.
Data comes back in the structure you defined. No post-processing after the node.
Configure an AI task on easybits and call it instantly in n8n.
Every task runs through the same easybits node in n8n. What changes is how you configure it on the platform, not what you wire on your canvas.
Answers a question using only the material you pass in, and returns the passage it drew on.
Returns answer · passage · confidence
Sorts a ticket, an email or a message into labels you define, with a confidence score.
Returns label · confidence · alternatives[]
Pulls named fields from a receipt, invoice or CV into clean JSON.
Returns fields{} · confidence · source_page
Turns a recording or a voice note into text, with timings and the language it detected.
Returns transcript · segments[] · language
Condenses a long document or thread down to the points that matter.
Returns summary · key_points[] · length
Reads text back as an audio file, in the voice and pace you pick.
Returns audio_url · duration · voice
Writes new copy from your brief, in the format and tone you set.
Returns text · format · finish_reason
Moves text between languages while leaving names, codes and formatting alone.
Returns translation · source_language · kept_terms[]
Reads a photo or a screenshot and returns what is in it, along with any text it contains.
Returns description · objects[] · text_found
Finds names, addresses and account numbers and masks or removes them.
Returns redacted_text · found[] · categories[]
Answers a question using only the material you pass in, and returns the passage it drew on.
ReturnsYou never have to compare models. Each task already has one selected, and two alternatives we recommend sit beside it.
We keep that list current. When a model is retired or a better one appears, we swap it and your workflow keeps running.
