Overview
Recall.ai provides APIs and SDKs to capture, transcribe, and surface meeting recordings and metadata across video conferencing (Zoom, Google Meet, Microsoft Teams), Slack Huddles, and in-person/phone via SDKs. Core capabilities include a Meeting Bot API (white-label bots that join meetings, record audio/video, produce transcripts, speaker timelines, participant metadata and media files), a Desktop Recording SDK for macOS and Windows (on-device recorder that detects meetings and uploads media without a bot joining), a planned Mobile SDK for phone/in-person recording, real-time and async transcription modes (prioritize_accuracy and prioritize_low_latency), and streaming/output media features for low-latency audio/video and interactive meeting agents. Authentication is via API keys managed in the Recall dashboard and passed in the HTTP Authorization header. Pricing includes Pay As You Go ingestion ($0.70 per hour of recording, first 5 hours free on signup), storage retention charges ($0.05 per hour beyond the default free window), transcription ($0.15 per hour for both real-time and async), and custom Launch/Enterprise plans for enterprise needs (SSO, HIPAA BAA, uptime guarantees, white-glove Slack, custom contracts, volume discounts). Storage defaults and retention behaviors vary by account creation date (docs note accounts created after June 12, 2025 may differ); default free retention is typically 7 days per recording unless configured otherwise. Security posture and compliance claims include SOC 2, ISO 27001, GDPR, CCPA, and HIPAA (blog announcement); data at rest is AES-256 encrypted and TLS is used in transit. Logs are retained for seven days even after media deletion per docs. Developer docs cover Bots, Desktop Recording, Recordings, Calendar v1/v2, Zoom OAuth, Slack Huddles, Billing, Meeting Direct Connect, and a getting-started hub. Enterprise/Launch pricing and exact SLA/contract terms require contacting sales. Source pages used are linked in external_links.
Key Features
Meeting Bot API
White-label bots join meetings to record audio/video, produce transcripts, speaker timelines, participant metadata (including participant emails in supported flows), and media files.
Desktop Recording SDK
On-device recorder for macOS and Windows that detects meetings and uploads media without requiring a bot to join the call (desktop upload flow via createDesktopSdkUpload and upload token).
Planned Mobile SDK
Mobile SDK for phone/in-person recording is planned/coming soon (mobile recording capability noted as forthcoming).
Real-time and Async Transcription
Built-in transcription priced at $0.15 per hour; modes include prioritize_accuracy (async/end-of-call) and prioritize_low_latency (real-time, 1–3s typical latency, language and feature restrictions).
Streaming and Interactive Media
Low-latency audio/video streaming into calls to support interactive meeting agents and streaming-only/zero-data retention options.
Storage and Retention Controls
Default free storage window (typically 7 days) with configurable retention (Timed or Forever); charges begin when a recording is marked done; deletion is permanent.
Who Can Use This Tool?
- developers:Build meeting capture, recording uploads, transcription, and integrations with conferencing platforms via APIs and SDKs.
- product teams:Add meeting recording, search, and transcript features to products and internal tools; enable real-time agents and streaming interactions.
- enterprises:Negotiate Launch/Enterprise plans for SSO, HIPAA BAA, uptime guarantees, custom contracts, and volume discounts.
Pricing Plans
Usage-based plan: recording ingestion at $0.70/hour, first 5 hours free on signup; storage retention billed $0.05/hour after default free window.
- ✓Recording ingestion: $0.70 per recording hour (first 5 hours free on signup)
- ✓Transcription: $0.15 per hour (real-time and async)
- ✓Default free storage per recording: typically 7 days; retention beyond free window billed $0.05 per hour (prorated per second)
- ✓Billing is per-second
- ✓Limits: 500-hour monthly limit on Pay As You Go; 2-hour per-recording limit
- ✓Support channels: email, in-dashboard chat, Ask AI in docs (for Pay As You Go)
Custom enterprise plans with negotiated pricing, SLAs, and enterprise features.
- ✓Custom pricing and volume discounts
- ✓White-glove Slack support and dedicated onboarding
- ✓SSO and enterprise identity integrations
- ✓Uptime guarantees and contractual SLAs
- ✓HIPAA BAA available (per announcement)
- ✓Custom contracts and negotiated terms via sales
Built-in Recall.ai transcription service: $0.15 per hour for both real-time and async transcription.
- ✓$0.15 per hour (same price for real-time and async)
- ✓Modes: prioritize_accuracy (async, language detection) and prioritize_low_latency (real-time, 1–3s latency typical)
- ✓Transcripts delivered via webhooks or download URL in recordings.media_shortcuts.transcript
Pros & Cons
✓ Pros
- ✓Comprehensive meeting capture across common conferencing platforms and Slack
- ✓Built-in transcription with same per-hour price for real-time and async ($0.15/hour)
- ✓Desktop SDK that records on-device without bot presence
- ✓Flexible delivery options (webhooks, download URLs) and streaming/interactive features
- ✓Enterprise features and compliance options (SSO, HIPAA BAA, SOC 2, ISO 27001)
✗ Cons
- ✗Enterprise/Launch pricing, SLAs, and custom terms require sales negotiation (no public fixed enterprise rates)
- ✗Pay As You Go has documented limits (500-hour monthly limit, 2-hour per-recording limit) that may constrain some use cases
- ✗Retention charges begin when recording is marked done; deletion is permanent and logs retained seven days
- ✗Some documentation and feature bullets are extensive and automated extraction may truncate details — targeted parsing may be needed for machine-readable exports
Compare with Alternatives
| Feature | Recall.ai | Krisp | Circleback.ai |
|---|---|---|---|
| Pricing | N/A | $8/month | $20.83/month |
| Rating | 8.2/10 | 8.1/10 | 8.2/10 |
| Capture Modes | Yes | Partial | Partial |
| SDK Depth | Desktop SDK mobile SDK planned API first | Developer SDKs and audio models | API first limited SDKs |
| Real-time Streaming | Yes | Yes | No |
| Audio Processing | Streaming audio capture limited advanced processing | Advanced noise cancellation and audio enhancement | Basic audio processing for transcription |
| Transcription & NLP | Real time and async transcription metadata surfacing | Real time transcription and summaries | Automatic transcription notes and action NLP |
| Action-item Extraction | No | Partial | Yes |
| Storage & Retention | Yes | Partial | Yes |
| Security & Compliance | Yes | Yes | Yes |
Related Articles (4)
A comprehensive comparison of 8 Zoom transcript methods, from Cloud Recording to Recall.ai, covering real-time access, diarization, and costs.
A practical tutorial for integrating Recall.ai's Desktop Recording SDK to detect, record, transcribe, and retrieve meetings in Electron apps.
An all-in-one Voice AI platform offering real-time STT, TTS, and agent orchestration for enterprises.
A marketing-focused overview of Recall.ai’s API and SDKs for reliable, scalable meeting recording and transcription.
