Skip to content
Neattrim

Neattrim API and developer docs

View as Markdown · Open in Claude Desktop

Run video, audio, image and AI tools from your own code with one API key. Every account can use the API, free accounts included, within its allowance.

Get an API key API reference · Use from an AI assistant · Pricing

Your first call

Estimate what a one-minute video compression would draw from your allowance. The call works with a test key or a live key, needs no file, and draws nothing.

A free account includes 1 hour of subtitle time and 5 minutes of voice and AI time each month. The website, the API and MCP share that allowance. Test keys draw nothing.

curl -X POST https://api.neattrim.com/v1/estimate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"tool":"compress_video","duration_seconds":60}'

Tools and what they draw

Cloud tools are called by REST and by the hosted MCP server, and draw from your allowance as shown. Tools that run on your computer have no cloud endpoint: use them through the local MCP server or the command line. Usage is counted in hours and minutes.

ToolWhere it runsDraws
Probe mediaCloud: REST API and MCPDoes not draw from the allowance.
Trim videoCloud: REST API and MCP6 seconds of voice and AI time for each job.1 minute of voice and AI time for each minute of output.
Compress videoCloud: REST API and MCP1 minute of voice and AI time for each minute of output.
Convert mediaCloud: REST API and MCP6 seconds of voice and AI time for each job.1 minute of voice and AI time for each minute of output.
Edit videoCloud: REST API and MCP1 minute of voice and AI time for each minute of output.6 seconds of voice and AI time for each job.
Make GIFCloud: REST API and MCP1 minute of voice and AI time for each minute of output.
Make memeCloud: REST API and MCP6 seconds of voice and AI time for each image.1 minute of voice and AI time for each minute of output.
Edit audioCloud: REST API and MCP1 minute of voice and AI time for each minute of output.
Edit imageCloud: REST API and MCP6 seconds of voice and AI time for each image.
Convert subtitlesCloud: REST API and MCPDoes not draw from the allowance.
Transcribe mediaCloud: REST API and MCP1 minute of subtitle time for each minute of audio.2 seconds of voice and AI time for each minute of audio.
Translate subtitlesCloud: REST API and MCP1 minute of subtitle time for each minute of video, for each language.
Review translationCloud: REST API and MCP1 minute of subtitle time for each minute of subtitles.
Burn captionsCloud: REST API and MCP1 minute of voice and AI time for each minute of output.
Subtitle videoCloud: REST API and MCP1 minute of subtitle time for each minute of audio.1 minute of subtitle time for each minute of video, for each language.1 minute of voice and AI time for each minute of output.
Make notesCloud: REST API and MCP1 minute of subtitle time for each minute of audio.15 seconds of voice and AI time for each minute of source.1 minute of voice and AI time for each job.
Clean audioCloud: REST API and MCP1 minute of voice and AI time for each minute of audio.
Remove musicYour computer: local MCP server and CLIDoes not draw from the allowance.
Separate vocalsYour computer: local MCP server and CLIDoes not draw from the allowance.
Remove image backgroundCloud: REST API and MCP6 seconds of voice and AI time for each image.
Make passport photoCloud: REST API and MCP6 seconds of voice and AI time for each image.
Clean up talking videoCloud: REST API and MCP1 minute of subtitle time for each minute of audio.2 seconds of voice and AI time for each minute of audio.1 minute of voice and AI time for each minute of audio.1 minute of voice and AI time for each job.1 minute of voice and AI time for each minute of output.
Process podcastCloud: REST API and MCP1 minute of subtitle time for each minute of audio.2 seconds of voice and AI time for each minute of audio.15 seconds of voice and AI time for each minute of source.1 minute of voice and AI time for each job.1 minute of voice and AI time for each minute of output.
Smart edit videoCloud: REST API and MCP1 minute of subtitle time for each minute of audio.1 minute of voice and AI time for each job.1 minute of voice and AI time for each minute of output.
Make shortsCloud: REST API and MCP1 minute of subtitle time for each minute of audio.15 seconds of voice and AI time for each minute of source.1 minute of voice and AI time for each minute of output.
Reframe videoCloud: REST API and MCP1 minute of voice and AI time for each minute of output.
Text to speechCloud: REST API and MCP1 minute of voice and AI time for each minute of speech.2 minutes of voice and AI time for each minute of speech (Pro plans only)
Dub videoCloud: REST API and MCP1 minute of voice and AI time for each minute of video.2 minutes of voice and AI time for each minute of video (Pro plans only)
Text to videoCloud: REST API and MCP1 minute of voice and AI time for each job.1 minute of voice and AI time for each minute of speech.1 minute of subtitle time for each minute of audio.1 minute of voice and AI time for each minute of output.
Generate reelCloud: REST API and MCP1 minute of voice and AI time for each job.1 minute of voice and AI time for each minute of speech.1 minute of voice and AI time for each minute of output.

API access for every account

No plan is needed to use the API or MCP. The allowance of your plan is what limits you.

PlanSubtitle time a monthVoice and AI time a month
Free1 hour5 minutes
Plus10 hours20 minutes
Pro50 hours2 hours

Webhooks

Get a signed event when a job ends, so you do not need to poll. Events follow the Standard Webhooks layout, and the SDKs check the signature for you.

Read about webhooks

Batches

Run one tool over many files in one request: up to 20 files on Free, 100 on Plus and 500 on Pro.

Read about jobs and batches

MCP and AI assistants

Connect Claude, ChatGPT and coding agents to the same tools, or run the local server on your own computer.

Connect with MCP

Rate limits and errors

A free account may send 60 requests a minute. Every error has a stable code and a fix, and every response shows how many requests you have left.

Read usage and limits

Status

Check whether the API is up before you debug.

Open the status page

More in the docs

Use from an AI assistant