Mattering Automations AssemblyAI integration
Connect AssemblyAI to Mattering Automations: submit audio and video for transcription by URL, poll jobs to completion, export sentences, paragraphs, and SRT or VTT subtitles, search transcripts for keywords, fetch PII-redacted audio, and run summaries, action items, translations, and LLM chat over the results, authenticated with one API key.
The Mattering Automations AssemblyAI integration adds AssemblyAI's speech AI to your automation canvas: submit audio or video files for transcription by URL, poll jobs through to completion, read finished transcripts as sentences, paragraphs, or SRT and VTT subtitles, search them for specific words and phrases, fetch PII-redacted audio, and analyze the results with summaries, action items, translations, and chat-style questions routed through AssemblyAI's LLM Gateway.
The credential is the account's API key, copied from the AssemblyAI dashboard. One key covers both surfaces this integration uses, the speech-to-text API and LLM Gateway, and every AssemblyAI account has API access from signup with $50 in free credits and no credit card required, so the whole integration can be set up and tested on a brand-new account. This guide walks through each step and ends with the full node reference.
The safety modeling is stated on the nodes rather than buried. Of the 14 nodes, seven read and seven act, and exactly one carries a destructive flag: AssemblyAI: Delete Transcript, which permanently removes a transcript and everything derived from it. Two surfaces are deliberately absent: streaming speech-to-text runs over WebSocket connections, which do not fit a canvas built on request-and-response steps, and raw file upload is not shipped because the submit node hands AssemblyAI a URL and AssemblyAI downloads the file itself. LeMUR, AssemblyAI's earlier LLM framework, was retired on March 31, 2026, so the analysis nodes here target its replacement, LLM Gateway.
Before you start
- An AssemblyAI account. Signup is free, includes $50 in credits with no credit card required, and API access is included from the start, so a new account can exercise every node in this integration.
- Audio or video reachable by URL. The submit node passes AssemblyAI a URL and AssemblyAI downloads the file itself, so recordings need to live at a public or signed URL (cloud storage links work well). There is no raw file upload from the canvas.
- A Mattering workspace with access to the Automations app.
- For PII-redacted audio, plan the submission: the redacted file only exists for transcripts submitted with audio redaction enabled, and it cannot be produced after the fact for a transcript submitted without it.
Get your AssemblyAI API key
Mattering authenticates to AssemblyAI with the account's API key. There is exactly one credential to manage: the same key works for the speech-to-text API and for LLM Gateway. One detail this integration absorbs for you: AssemblyAI expects the raw key in the Authorization header with no Bearer prefix, which is a common stumbling point in hand-rolled integrations. Mattering formats the header itself, so all you handle is the key.
Connect AssemblyAI to Mattering
Verify it works
After connecting, open any automation canvas and check the palette: an AssemblyAI section should list the query and action nodes. Drop AssemblyAI: List Transcripts onto the canvas and run a test step. It is the integration's cheapest health check: free, read-only, and valid on every account, returning the account's transcript jobs newest first with each job's ID, status, and source audio URL. An empty list on a new account is a success, not a failure.
For an end-to-end proof, run AssemblyAI: Submit Transcript with the URL of a short public audio file. It returns immediately with the new transcript's ID and a status of queued. Re-run AssemblyAI: Get Transcript with that ID until the status reads completed, at which point the full text arrives in the output along with word-level timings and confidences. A short test clip costs a fraction of a cent against the signup credits, and AssemblyAI: Delete Transcript cleans it up afterwards if you want the account spotless.
Node reference
14 nodes, generated from the connector itself so this reference always matches what ships.
Read and query nodes 7
AssemblyAI: Get Redacted Audio
assemblyai_get_redacted_audio
Fetch the download URL for a transcript's PII-redacted audio file (the version with sensitive speech beeped out). Only exists when the transcript was submitted with redact_pii_audio true, and the returned URL expires 24 hours after generation, so download it promptly downstream.
AssemblyAI: Get Subtitles
assemblyai_get_subtitles
Export a completed transcript as subtitle text in SRT or VTT format, returned verbatim at data.text (capped at 200,000 characters with a truncation flag). Pipe it into a file step or a video pipeline downstream.
AssemblyAI: Get Transcript
assemblyai_get_transcript
Fetch a transcript job by id: its status plus, once completed, the full text, word-level timings and confidences, and any enabled intelligence results (highlights, sentiment, entities, content safety). Status moves queued to processing to completed or error, so this is the polling read after AssemblyAI: Submit Transcript: re-run it until data.status is completed, then branch on the result.
AssemblyAI: Get Transcript Paragraphs
assemblyai_get_paragraphs
Fetch a completed transcript semantically segmented into reader-friendly paragraphs, each with start/end times, a confidence score and word-level detail. The right shape for publishing a transcript as readable text.
AssemblyAI: Get Transcript Sentences
assemblyai_get_sentences
Fetch a completed transcript segmented into sentences, each with start/end times in milliseconds, a confidence score, the speaker (when diarization was enabled) and word-level detail. The natural unit for clipping, quoting and sentence-level analysis downstream.
AssemblyAI: List Transcripts
assemblyai_list_transcripts
List the account's transcript jobs, newest first, with each job's id, status, source audio URL and timestamps. Filter by status or creation date; page with before_id/after_id cursors (data.page_details carries ready-made prev_url and next_url). The transcript IDs from here feed every other AssemblyAI node, and this is also the cheapest credential health check to run first in any AssemblyAI automation.
AssemblyAI: Word Search
assemblyai_word_search
Search a completed transcript for words, numbers or short phrases (up to five words each). Returns each match with its occurrence count, millisecond timestamps and word indexes, so an automation can jump straight to the moments that matter (competitor mentions, pricing talk, names).
Action nodes 7
AssemblyAI: Delete Transcript
assemblyai_delete_transcript
DESTRUCTIVE: permanently remove a transcript's data from AssemblyAI. The transcript text, source audio URL and all derived results are irrecoverably deleted; the id remains but returns placeholder values. Capture anything you need with AssemblyAI: Get Transcript first.
AssemblyAI: Extract Action Items
assemblyai_transcript_action_items
Extract action items from a completed transcript: each item comes with the supporting quote and its millisecond timestamp, at data.speech_understanding.response.action_items.items. The meeting-recording to task-list automation in one node. Billed per request.
AssemblyAI: LLM Chat Completion
assemblyai_llm_chat
Run a prompt through AssemblyAI's LLM Gateway (the LeMUR replacement): an OpenAI-compatible chat completion routed to Claude, GPT, Gemini and other models. Pass a transcript_id and AssemblyAI injects that transcript's text into the conversation server side, so you can ask questions about a call without templating the whole transcript through the canvas. The answer arrives at data.choices[0].message.content. Billed per token.
AssemblyAI: Speech Understanding Request
assemblyai_speech_understanding
Run any speech understanding request on a completed transcript with the raw request object: speaker identification (name or role), custom formatting of dates/phone numbers/emails, or the summarization, action items and translation shapes the dedicated nodes cover. The generic escape hatch when a new understanding capability ships before its own node does. Billed per request.
AssemblyAI: Submit Transcript
assemblyai_submit_transcript
Submit an audio or video file for transcription by URL. AssemblyAI downloads the file itself, so no upload step is needed: returns immediately with the new transcript's id and status queued. Poll AssemblyAI: Get Transcript with that id until status is completed (or set webhook_url and let AssemblyAI call back when done). Billed per audio duration. Reversible with AssemblyAI: Delete Transcript.
AssemblyAI: Summarize Transcript
assemblyai_summarize_transcript
Summarize a completed transcript with AssemblyAI's speech understanding service (the LeMUR summary replacement). Returns timestamped summary sections with headlines at data.speech_understanding.response.summarization.summary. Billed per request.
AssemblyAI: Translate Transcript
assemblyai_translate_transcript
Translate a completed transcript into one or more target languages. Full translations arrive at data.translated_texts keyed by language code, with per-utterance translations alongside the original text. Billed per request.
Using the integration
The submit-then-poll pattern
Transcription is asynchronous, and the nodes are shaped for it. AssemblyAI: Submit Transcript returns immediately with the job's ID and a status of queued, and the job moves through processing to completed, or to error with the reason attached. Re-run AssemblyAI: Get Transcript with that ID until the status is completed, then branch on the result. Alternatively, set a webhook URL at submit time and AssemblyAI calls your workflow back when the job finishes, which removes the polling loop entirely. Every derived read, from sentences to word search, only has data once the transcript is completed.
From AI assistants over MCP
Every node speaks MCP, the open protocol for connecting AI assistants to tools, so Claude, ChatGPT, and Matter can drive the same surface. An agent can submit a recording, poll it to completion, pull the action items with AssemblyAI: Extract Action Items, and open a task for each one. The read nodes are the safest surface to hand an agent; keep AssemblyAI: Delete Transcript behind a condition or an approval step.
Readable exports: sentences, paragraphs, subtitles, keywords
A raw transcript is one long string; these nodes give it useful shapes. AssemblyAI: Get Transcript Sentences segments it into sentences with start and end times and, when diarization was enabled, the speaker, which is the natural unit for clipping and quoting. AssemblyAI: Get Transcript Paragraphs returns reader-friendly paragraphs for publishing. AssemblyAI: Get Subtitles exports SRT or VTT text directly in the node output, capped at 200,000 characters with a truncation flag, ready for a file step. And AssemblyAI: Word Search finds words, numbers, or phrases of up to five words, returning each match with its occurrence count and millisecond timestamps.
Analysis on the canvas: the LLM Gateway nodes
These nodes replace the retired LeMUR surface with its successor. AssemblyAI: Summarize Transcript returns timestamped summary sections with headlines, AssemblyAI: Extract Action Items returns each follow-up with its supporting quote and timestamp, and AssemblyAI: Translate Transcript produces full translations keyed by language code. AssemblyAI: LLM Chat Completion is the open-ended one: an OpenAI-compatible chat completion routed to Claude, GPT, Gemini, and other models, and passing a transcript ID makes AssemblyAI inject that transcript's text server side, so you can ask questions about a call without templating the whole transcript through the canvas. When a capability has no dedicated node, such as speaker identification or custom formatting of dates and numbers, AssemblyAI: Speech Understanding Request takes the raw request object.
PII redaction with a downloadable redacted file
Submit a recording with PII redaction and audio redaction enabled, and AssemblyAI produces a version of the file with the sensitive speech beeped out. AssemblyAI: Get Redacted Audio returns the download URL for that file once it is ready. The URL expires 24 hours after it is generated, so the reliable pattern is to download or forward the file in the same workflow run, not to store the link for later. This is the shape for call recordings that need to be shareable with people who should not hear card numbers or health details.
The one destructive node, gated
AssemblyAI: Delete Transcript permanently removes a transcript's text, its stored audio URL, and all derived results from AssemblyAI; the ID remains but returns placeholder values, and there is no undo. It exists for good reasons, from data retention policies to cleaning up test jobs, and the pattern for using it safely is the same as for any destructive node: capture what you need with AssemblyAI: Get Transcript first, and put the delete behind a condition, an approval task, or a manual run rather than a bare trigger.
Troubleshooting
Test Connection fails
A transcript sits at queued or processing
The transcript status is error
Sentences, paragraphs, subtitles, or word search return nothing
AssemblyAI: Get Redacted Audio fails or the link stops working
An older LeMUR workflow or tutorial does not match these nodes
Requests start failing during a heavy sweep
Related
The AssemblyAI MCP
What connecting AssemblyAI unlocks for AI agents and automations, with common uses and FAQs.
More guides
All Automations documentation
Every integration guide, plus how connectors, credentials, and workflows fit together.
AssemblyAI is a trademark of its respective owner. Mattering uses platform names to identify the services these integrations connect to; no endorsement or affiliation is implied.