Use Speak AI with Zapier
Connect Speak AI to 7,000+ apps with zero code. Send transcripts to your CRM. Push AI summaries into Notion or Slack. Auto-tag recordings by Magic Prompt result. The Speak Zapier app ships triggers and actions for transcripts, sentiment, recordings, exports, and Magic Prompt history.
Yapabilecekleriniz
Zapier lets non-developers connect Speak to anything. Sales ops route every analyzed call into HubSpot. Research teams push verbatim quotes from the Speak recorder into Notion. Support managers fire Slack alerts on negative sentiment recordings. All with zero code.
Trigger Zaps on every analyzed recording
Speak'in New Media Analyzed trigger fires when transcription and AI analysis complete. Zap it to your CRM, BI tool, or messaging app to surface the recording the moment it lands. No webhook code, no polling logic, no signature verification on your side.
Upload media into Speak from any Zap
Speak'in Upload Media action accepts a URL or file from the previous Zap step. Pull a recording from Twilio, Zoom Cloud Storage, Google Drive, Dropbox, or Loom and forward it to Speak in a single Zap step. Speak handles transcription and analysis automatically.
Run Magic Prompts as Zap actions
Use Speak’s Run Magic Prompt action inside a Zap to score a transcript against a custom rubric, extract decisions, or pull verbatim quotes. The next Zap step writes the structured answer into Notion, Airtable, your CRM, or anywhere else.
Export transcripts to any file format
Speak'in Export Transcript action delivers SRT, VTT, DOCX, PDF, JSON, or plain text with optional speaker names, timestamps, and PII redaction. Drop the export into Drive, Dropbox, S3, or attach it to a Slack message in the same Zap.
3 Adımda Kurulum
The Speak Zapier app is free on every Zapier plan. Connect once with your Speak API key, then build any workflow with the visual Zap editor.
Speak AI’ye kaydolun
Ücretsiz hesap oluştur app.speakai.co7 günlük deneme sürümü alırsınız ve tam erişim sunar. Kredi kartı gerekmez. İçeri girdikten sonra şuraya gidin Ayarlar > API and copy your API key. You will paste it into Zapier in step 2.
Install the Speak AI Zapier app
Ziyaret etmek zapier.com/apps/speak-ai/integrations. Click Connect Speak AI. Paste your Speak API key when prompted. Zapier validates the key against Speak’s API and stores the connection for reuse across Zaps.
Triggers fire when something new happens in Speak (new analyzed recording, new Magic Prompt result, new recording uploaded). Actions do something in Speak (upload media, run a Magic Prompt, export a transcript). Mix and match across Zap steps to build any pipeline.
Pick the second app in your Zap (HubSpot, Notion, Slack, Airtable, Google Drive, Sheets, anything Zapier supports). Map fields from the Speak step into the destination’s required fields. Zapier shows a sample payload from a real Speak record so you see exactly what is available.
Already have recordings in Speak? Connect Claude Desktop with npx @speakai/mcp-server init, or add the MCP URL to Claude.ai or ChatGPT. Use Zapier for app-to-app routing and MCP for natural-language search across the same library.
Turn the Zap on and test
Zapier walks you through a test run with real data. Pick a recent Speak recording, fire the Zap once, verify the destination got what you expected. Save and turn the Zap on. From that point, every new event fires the Zap automatically.
Gerçek iş akışları, gerçek sonuçlar
Four high-leverage Zaps Speak customers ship in production. Each uses Speak as a step in a multi-app pipeline. Pick one, copy the structure, customize for your tools.
Auto-log every analyzed call as a CRM activity
Three-step Zap. Speak’s New Media Analyzed trigger fires when transcription completes. The HubSpot Create Meeting Engagement action logs the call against the matching contact. The Slack Send Channel Message action notifies the deal channel. Same pattern works for Salesforce Tasks, Pipedrive Activities, Zoho Calls.
# Step 1: Trigger
# App: Speak AI
# Event: New Media Analyzed
# Output: mediaId, name, summary, sentiment, transcript, duration,
# tags, mediaUrl, createdAt
#
# Step 2: Action - lookup contact (optional)
# App: HubSpot
# Event: Find Contact
# Search: email matches {{first_attendee_email}}
#
# Step 3: Action - log the call
# App: HubSpot
# Event: Create Meeting Engagement
# Fields:
# hs_meeting_title: {{2.name}}
# hs_meeting_body: AI Summary: {{2.summary}}nnOpen in Speak: https://app.speakai.co/media/{{2.mediaId}}
# hs_meeting_external_url: https://app.speakai.co/media/{{2.mediaId}}
# hs_timestamp: {{2.createdAt}}
# contactIds: {{1.id}} # from step 2
#
# Step 4: Action - notify the team (optional)
# App: Slack
# Event: Send Channel Message
# Channel: #sales-pipeline
# Text: :phone: New call from {{2.name}} ({{2.sentiment.label}}). View: {{2.mediaUrl}}// Speak's New Media Analyzed trigger backs onto GET /v1/zapier/insights
// Zapier polls this endpoint and emits one Zap per new analyzed media.
{
"mediaId": "14c8dd9c0a89",
"name": "Acme Corp - Discovery",
"mediaType": "video",
"duration": { "inSecond": 1265, "start": "00:00:01.280", "end": "00:21:00.995" },
"sourceLanguage": "en-US",
"state": "processed",
"tags": ["sales-call", "discovery"],
"sentiment": [{
"document": {
"Compound": 25.10,
"Negative": 3.22,
"Neutral": 41.66,
"Positive": 55.10
}
}],
"insight": {
"transcript": [...],
"speakers": [...]
},
"mediaUrl": "https://app.speakai.co/media/14c8dd9c0a89",
"createdAt": "2026-05-06T17:22:51.269Z"
}For sentiment-aware routing, add a Filter step between trigger and action: {{sentiment.document.Compound}} less than -10 routes to a manager-attention channel; positive routes to a celebration channel.
Auto-transcribe every recording dropped into a folder
Two-step Zap. Cloud storage New File in Folder trigger fires when a new audio or video file lands. Speak’s Upload Media action ingests it via URL. Speak handles transcription, sentiment, and AI analysis. Optional third step: notify Slack when processing completes.
# Step 1: Trigger
# App: Google Drive (or Dropbox, OneDrive, Box)
# Event: New File in Folder
# Folder: /Research/Interviews/2026-Q3
#
# Step 2: Action
# App: Speak AI
# Event: Upload Media
# Fields:
# name: {{1.name}}
# url: {{1.webContentLink}} # or webViewLink
# mediaType: audio (or video)
# sourceLanguage: en-US
# tags: drive-import,research
# folderId: (your Speak folder ID, optional)
#
# Step 3: Action (optional)
# App: Slack
# Event: Send Channel Message
# Channel: #research-intake
# Text: :file_folder: New recording uploaded to Speak: {{2.name}}// Speak's Upload Media action backs onto POST /v1/media/upload
// Required: name, url
// Optional: mediaType, description, sourceLanguage, tags, folderId,
// callbackUrl (overrides webhook config), isTranscriptionOnly, fields
{
"name": "Customer interview - Acme - 2026-05-08",
"url": "https://drive.google.com/uc?id=1AbCd...",
"mediaType": "audio",
"sourceLanguage": "en-US",
"tags": "drive-import,interview,research",
"folderId": "67e9c4d2a8b1f3045e2c1a09"
}
// On success, Zap step 2 emits the new media's ID and metadata for
// downstream Zap steps to use.For Drive folders containing both audio and video, add a Path step that branches on file extension. Set mediaType: video for mp4/mov, mediaType: audio for mp3/m4a/wav.
Generate a Notion-ready writeup from every recording automatically
Three-step Zap. Speak’s New Media Analyzed fires. Speak’s Run Magic Prompt action runs a structured prompt against the new transcript (“Generate a Notion writeup with: ## Summary, ## Key decisions, ## Action items, ## Verbatim quotes”). The Notion Create Database Item action saves the result as a new page in your research database.
# Step 1: Trigger
# App: Speak AI
# Event: New Media Analyzed
#
# Step 2: Action - run a structured prompt
# App: Speak AI
# Event: Run Magic Prompt
# Fields:
# mediaId: {{1.mediaId}}
# prompt: "Generate a Notion writeup with: ## Summary,
# ## Key decisions, ## Action items,
# ## Verbatim quotes (3 of them with timestamps)"
# pollUntilComplete: true (Zapier waits for the answer)
#
# Step 3: Action - save to Notion
# App: Notion
# Event: Create Database Item
# Database: Research Transcripts
# Properties:
# Title: {{1.name}}
# Date: {{1.createdAt}}
# Sentiment: {{1.sentiment.label}}
# Speak URL: https://app.speakai.co/media/{{1.mediaId}}
# Page Body: {{2.answer}} # markdown from Magic Prompt// Speak's Run Magic Prompt action backs onto POST /v1/prompt/ + polling
// /v1/prompt/messages until the answer completes (typical: 2-3 seconds).
//
// Zapier handles the fire-and-poll loop transparently; your Zap just
// gets the final answer as a step output.
{
"mediaIds": ["14c8dd9c0a89"],
"prompt": "Extract decisions, owners, and deadlines from this transcript.",
"isStream": false,
"isIndividualPrompt": true
}
// Step output:
{
"messageId": "9929d50c45ea",
"state": "completed",
"prompt": "Extract decisions, owners, and deadlines...",
"answer": "## Decisionsn- Acme will sign by EOQn- Pricing: $X/seatn...",
"totalCharacters": 53104,
"cost": 0.45
}Save the prompt once in Speak, reference it by promptId in the Zap, and you can update the synthesis logic without touching the Zap. New prompt fires on every future recording.
Pair Zapier with MCP for both routes into Speak
Zapier handles app-to-app routing. MCP handles natural-language search across your Speak library. Most teams use both. Zapier auto-pushes structured fields into your CRM and Notion. MCP lets analysts ask questions across thousands of recordings without leaving Claude or ChatGPT.
# Claude Desktop / Claude Code (kurulumunuzu otomatik olarak algılar)
npx @speakai/mcp-server init
# İstendiğinde Speak API anahtarını yapıştırın. Kurulum yaklaşık 2 dakika sürer.# Claude.ai (web) and ChatGPT MCP connector
# Settings > Integrations > Add MCP Server
# Remote URL: https://api.speakai.co/v1/mcp
# Auth header: x-speakai-key: YOUR_SPEAK_API_KEY
# Verify the connection responds:
curl -s https://api.speakai.co/v1/mcp
-H "x-speakai-key: $SPEAK_API_KEY"
-H "Accept: application/json"2. When to use Zapier vs MCP:
- Zapier: every new recording fires a fixed pipeline (push to CRM, append to Notion, alert Slack). Same fields, same destinations, every time.
- MCP: analyst asks ad-hoc questions (“which calls last week mentioned a competitor by name?”). Different question every time, no fixed pipeline.
- Zapier: connecting Speak to a non-Anthropic app (HubSpot, Notion, Slack, Airtable, Sheets, Drive, Dropbox, etc).
- MCP: working inside Claude or ChatGPT and wanting Speak data available alongside the rest of your workspace context.
- Her ikisi: Zapier for the routing layer, MCP for the analysis layer. Use them together; they read the same Speak workspace.
Claude.ai, Claude Desktop, Claude Code ve ChatGPT MCP bağlayıcılarıyla çalışır. MCP sunucusunu görüntüle →
Why Speak AI + Zapier
Speak AI provides the analysis layer. Zapier provides the routing layer to 7,000+ apps. The combination is how a 5-person sales team and a 5,000-person enterprise both auto-log every conversation into the systems where work happens, without writing or maintaining custom integration code.
Free on every Zapier plan
The Speak AI Zapier app costs nothing. It runs on Zapier’s free tier and on every paid plan. Most production setups fit comfortably under Zapier’s free or Starter tier task volumes since each new media event maps to a single Zap run.
Triggers and actions, not just webhooks
Zapier exposes Speak as a first-class integration: triggers on the left side of any Zap, actions on the right. Build multi-step workflows that pull from a CRM, push into Speak, transcribe, run a Magic Prompt, then write the result back into Notion or Sheets. All visual, all no-code.
Magic Prompt as a Zap step
Magic Prompt is exposed as a Zapier action. Run a structured analysis prompt as a step in any Zap, then route the structured answer to the next app. No model-handling code, no async polling logic, no API key management beyond the initial Zap connection.
One workspace, every input source
Speak ingests Zoom, Teams, Meet, Webex, Twilio, Loom, Drive, Dropbox, podcast audio, embedded recorder submissions, and direct file uploads. Zapier connects any of those upstream sources to Speak in one Zap step, then routes the analysis downstream to anywhere you need it.
Takımlar en önemli çağrıları için Speak AI’ye güveniyor
4.9 G2'de
“Speak AI, nitel verileri nasıl işlediğimizi dönüştürmede müdür olmuştur. Transkripsiyon doğruluğu etkileyici ve NLP içgörüleri bize manuel analiz saatlerini kurtarıyor.”
Araştırma Müdürü | Danışmanlık Firması
“Otter.ai’den geçtik ve analiz derinliği başka bir seviyede. Sentiment puanlama, anahtar kelime çıkarma ve tema tespiti otomatik olarak gerçekleşiyor.”
Ürün Yöneticisi | SaaS Şirketi
“Tüm müşteri aramalarında arama yapabilmek ve belirli anları çekebilmek, destek ekibimiz için oyun kurallarını değiştiren bir özellik.”
CX Müdürü | Enterprise Tech
How to use Speak AI with Zapier for no-code automation
Zapier is the dominant no-code automation platform. With 7,000+ supported apps and a visual Zap editor, it lets a sales ops manager, research lead, or solo founder connect Speak to their full toolchain in minutes instead of weeks. Every analyzed call can fan out into a CRM activity, a Notion page, a Slack alert, an Airtable row, a Google Sheet, a Mailchimp list, or any combination of those, without writing or maintaining code.
Where Zapier fits in the Speak workflow
Speak handles ingestion, transcription, and AI analysis. Zapier handles routing. The handoff happens at two points: Zapier triggers react to events in Speak (new analyzed media, new Magic Prompt result, new recording from the Speak embedded recorder), and Zapier actions yap things in Speak (upload media by URL, run a Magic Prompt, export a transcript). The combination covers both sides of any workflow that needs Speak as a step.
Triggers exposed by the Speak Zapier app
- New Media Analyzed. Fires when transcription and AI analysis complete. Outputs the full media insight: summary, sentiment, transcript, speakers, tags, duration, mediaUrl. Most common trigger for CRM and notification Zaps.
- New Recording (Recorder). Fires when someone submits a recording through the Speak embedded recorder. Useful for routing user-submitted feedback or research interviews.
- New Magic Prompt Result. Fires when a Magic Prompt finishes. Output is the structured answer plus the prompt that produced it. Common in research synthesis pipelines.
Actions exposed by the Speak Zapier app
- Upload Media. Sends a URL or file to Speak’s
/v1/media/uploadendpoint. Speak transcribes and analyzes it. Most common action. - Run Magic Prompt. Fires a structured prompt against an existing media or folder. Zapier handles the fire-and-poll loop transparently; the Zap step output is the final answer.
- Export Transcript. Generates SRT, VTT, DOCX, PDF, JSON, or plain text export with optional speaker names, timestamps, and PII redaction. Drop the export into Drive, Dropbox, S3, or attach to a Slack message in the next step.
Authentication and rate limits
Connect Speak to Zapier with a single API key. Generate the key in Speak under Settings > API. Paste it into the Speak Zapier app on the first Zap that uses Speak; Zapier reuses the connection across every subsequent Zap. Speak’s API rate limits apply across all Zapier traffic. The Magic Prompt endpoint is rate-limited to 15 requests per minute per API key, which is plenty for typical Zap volumes (one new analyzed media every few minutes at most).
Zapier vs custom webhooks
For most teams, Zapier is the right starting point. It is faster to build, easier to maintain, and visible to non-engineers who can troubleshoot or extend the Zap themselves. Custom webhooks (Speak’s POST /v1/webhook için media.analyzed) are the right path when you need: sub-second routing latency, custom signing or signature verification, complex routing logic that exceeds Zap step limits, or workflows that touch internal systems Zapier does not support natively.
Role’a Göre Kullanım Örnekleri
Satış ve RevOps ekipleri use Zapier with Speak to auto-log every analyzed call as a CRM activity (HubSpot Meeting, Salesforce Task, Pipedrive Activity, Zoho Call). Sentiment-aware filters route negative-sentiment calls to a manager-attention channel. See Satış ekipleri için Speak AI.
Müşteri araştırması ekipleri chain Drive (or Dropbox) New File in Folder -> Speak Upload Media -> Speak Run Magic Prompt -> Notion Create Database Item. New interviews become structured Notion writeups automatically. See Nitel araştırmacılar için Speak AI.
Müşteri desteği ve CX ekipleri route Speak’s analyzed support calls into Zendesk, Freshdesk, or Help Scout as ticket comments with sentiment and key-topic tags. Cancellation language fires a separate Slack alert via a Zap filter step.
Product and operations teams turn the Speak embedded recorder into a feedback intake pipeline. Recorder submissions trigger Zaps that tag the feedback by product area, score it on a custom rubric, and route it into the right team’s intake doc.
Sıkça sorulan sorular
Does Speak AI have a Zapier app?
Yes. The official Speak AI Zapier app lives at zapier.com/apps/speak-ai/integrations. It is free to install on any Zapier plan, including the free tier. Connect once with your Speak API key and reuse the connection across every Zap.
What triggers and actions does the Speak Zapier app expose?
Triggers: New Media Analyzed (most common), New Recording (from embedded recorder), New Magic Prompt Result. Actions: Upload Media, Run Magic Prompt, Export Transcript (SRT, VTT, DOCX, PDF, JSON). All triggers and actions back onto Speak’s documented API endpoints.
How do I trigger a Zap when a Speak recording is analyzed?
Use the New Media Analyzed trigger. Pick it as step 1 of your Zap, paste your Speak API key, and Zapier polls Speak for new analyzed media. Each new record fires the Zap with the full insight payload (summary, sentiment, transcript, speakers, tags) available to subsequent Zap steps.
Can I run a Magic Prompt as a Zap action?
Yes. The Run Magic Prompt action fires a prompt against an existing media or folder, polls until the answer completes, and returns the structured answer to the Zap. Typical completion is 2 to 3 seconds for a 50,000-character transcript. Zapier handles the polling loop transparently.
What is the difference between the Speak Zapier app and Speak’s webhooks?
Zapier is no-code routing for app-to-app workflows. Speak’s webhooks (POST /v1/webhook için media.analyzed) are the lower-level path for custom server logic. Most teams start with Zapier; advanced setups with sub-second routing latency or complex branching use webhooks directly. Both can run side by side in the same Speak workspace.
Can I try Speak AI for free with Zapier?
Yes. The Speak 7-day trial includes credits for transcription, full API access, and the Zapier app. No credit card required. Build a Zap, run a test recording through it, validate the full pipeline before committing.
Start using Speak AI with Zapier today
83 analysis tools. 100+ languages. 5+ Zapier triggers and actions. 7,000+ destination apps. Same Speak workspace as your Zoom, Teams, Meet, Webex, and Twilio calls.
Browse the Speak Zapier app
Open Speak on Zapier. Start a Zap, pick a trigger, connect your destination app. Free on every Zapier plan.
API belgelerini görüntüleyin
Full reference for Speak’s Zapier-backing endpoints, the webhook event types, and the Magic Prompt API. Plus the official MCP server on NPM.
English 
Español 
Français 
Italiano 
العربية 
Português do Brasil 
Nederlands 
Deutsch 
עִבְרִית 
Русский 
日本語 
Polski 
Čeština 
Українська 
Slovenščina 
Svenska 
Ελληνικά 
Norsk bokmål 
简体中文 
Türkçe 
Català 
Magyar 
한국어 
Slovenčina 
Suomi