Riverside
Riverside is an AI-powered studio in the browser that records every participant locally in up to 4K, then edits, repurposes, live-streams and publishes the result. Podcasts, webinars, interviews and social clips all come out of one session.
What is Riverside?
Riverside is a browser-based production platform for people who make audio and video content, built by RiversideFM, Inc. and founded in 2020 by brothers Gideon and Nadav Keyson. Its defining idea is local recording: instead of capturing a compressed video call, Riverside records each participant on their own device and uploads the files to the cloud continuously while the session runs. The result is up to 4K video and uncompressed 48 kHz WAV audio for every speaker, on separate tracks, unaffected by a shaky connection and recoverable even if a browser crashes. Guests join with one link, with no account and no installation, and up to ten people (host plus nine guests) can be recorded, while extra viewers watch without being captured.
Editing happens in the same place. A transcript is generated automatically and the video follows whatever you do to the text, so cutting a sentence cuts the footage; a conventional timeline sits alongside it for finer work, and edits are non-destructive. On top of that sits a broad AI layer: Magic Clips finds shareable moments, Magic Audio cleans the sound, VideoDub regenerates speech in the speaker's own voice and lip-syncs it after you fix the transcript, and there are tools for eye contact, automatic layouts, filler and silence removal, generated B-roll, captions and translation into more than 30 languages. An AI Co-Creator agent turns a session into clips, captions, thumbnails, newsletters and articles from a chat prompt. Transcription is advertised in over 100 languages.
The platform then handles distribution. It live-streams in 1080p to several destinations at once, runs branded webinars with registration, reminders, polls, Q&A and HubSpot sync, and hosts podcasts and video with analytics, a creator site and social scheduling. Work is organised into studios, one per show. Business customers add producer controls, shared workspaces, roles, async recording, XML export to Premiere or Final Cut, Salesforce and Marketo, SSO, API access and SOC 2 Type II and ISO 27001 coverage. Mac, iOS, Android and iPad apps complete the set, and the site advertises a 4.8 rating on G2 from 1,582 reviews.
What it does
- Record every participant locally in up to 4K video and uncompressed 48 kHz audio, on separate tracks
- Edit the video by editing its transcript, or switch to a classic timeline, entirely in the browser
- Turn one recording into social clips, show notes, titles, chapters, thumbnails and blog posts with AI
- Go live in 1080p to YouTube, Facebook, LinkedIn, X, Twitch or any custom RTMP destination at once
- Run branded webinars with registration, reminders, Q&A, polls, call-ins and HubSpot lead sync
- Host and distribute a podcast to Spotify, Apple and YouTube with built-in analytics
- Translate and dub a recording into more than 30 languages, keeping the speaker's voice and lip-sync
When to use Riverside / When not to
A quick filter to help you decide if Riverside is the right fit.
When to use Riverside
- Podcasters and video creators who record remote guests and need broadcast-grade files rather than a compressed call recording
- Marketing and communications teams turning one session into clips, posts, newsletters and blog copy at scale
- Producers and media agencies who need separate tracks, producer controls, shared workspaces and role-based permissions
- B2B teams running lead-generation webinars with registration, Q&A, reminder emails and HubSpot sync
- Learning and development teams producing training, onboarding and internal communication videos without a studio
When not to use Riverside
- Teams looking for a day-to-day meeting or videoconferencing tool: Riverside is built for content production, not internal calls
- Large panel formats, since only ten people (host plus nine guests) can be captured on separate tracks
- Anyone expecting a usable free tier, as the free plan caps quality at 720p, stamps a watermark and allows a single two-hour multitrack session
- Developers wanting to build on top of the platform, because API access is sold with the Business plan but no public documentation exists
- Users under 18, and anyone who needs an interface in a language other than English or German
How to use Riverside
A typical end-to-end flow, from setup to results.
- Create an account from the homepage: the free plan needs no credit card
- Open a studio, which is the dedicated space for one show, channel or client
- Send guests the invitation link; they join in a browser with no account and nothing to install
- Use the lobby to check microphones, cameras and lighting before anyone goes live
- Record: each participant is captured locally while files upload to the cloud in the background
- Open the recording in the built-in editor, where a transcript has already been generated
- Cut and rearrange by editing the transcript, or refine on the timeline; nothing is destructive
- Apply AI where it helps: clip detection, audio clean-up, filler removal, captions, translation or B-roll
- Apply your Brand Kit so fonts, colours, logo, intro and outro carry across every export
- Publish straight to YouTube, Spotify or Apple, or download the raw tracks and, on Business, an XML timeline for Premiere or Final Cut
Pros & Cons
Pros
- Recording quality does not depend on anyone's internet connection, because capture happens on each device
- Separate tracks per participant make cross-talk removal, layout changes and professional post-production straightforward
- Guests join from one link with no account and no download, which removes most of the friction of remote interviews
- Capture, editing, repurposing, streaming, hosting and distribution live in a single tool instead of five
- Transcript-based editing makes video editing accessible to people who have never opened a video editor
- A permanent free plan and a 14-day trial on paid tiers let you evaluate the platform without a credit card
- Security and privacy posture is documented: SOC 2 Type II, ISO 27001, a published DPA and a named subprocessor list
Cons
- The privacy policy lists training Riverside's own AI and ML models among its purposes, and no opt-out mechanism is documented anywhere on the site
- The free plan is severely limited: 720p, a watermark, 44.1 kHz audio and a single two-hour multitrack allowance
- Five tiers with overlapping features make plan selection genuinely confusing; the site even offers a quiz to help you choose
- Separate-track downloads are capped monthly at 15, 20 or 25 hours depending on the plan
- The heaviest AI features consume a limited credit allowance, only 20 credits on Pro and Grow
- Collaborative editing, roles, SSO, Salesforce and API access are locked behind Business, whose price is not published
- API access is sold but no public developer documentation exists, and the interface itself is only available in English and German
Pricing & Plans
A permanent free plan is available and requires no credit card, although it is limited to 720p video, 44.1 kHz audio, a Riverside watermark and a single two-hour multitrack allowance. The cheapest paid tier is Pro at USD 24.00 per month when billed annually, which amounts to USD 288 per year; the same plan costs USD 29.00 per month without an annual commitment. Grow and Webinar follow at USD 34.00 and USD 79.00 per month on annual billing, and the Business tier is quoted on request. Annual billing saves up to 20 percent, and Pro, Grow and Webinar each open with a 14-day free trial.
- one 2-hour multitrack allowance
- 720p video
- 44.1 kHz audio
- Riverside watermark
- unlimited single-track recording and editing
- Magic Clips and unlimited video calls
- up to 4K video
- 48 kHz audio
- no watermark
- host teleprompter
- in-person multi-camera recording
- unlimited text-based editing and transcriptions
- AI editing and repurposing agent
- Magic Clips and show notes
- everything in Pro plus social scheduling
- newsletter
- creator website
- hosting for 2 shows
- video hosting
- AI thumbnails
- 1080p multistreaming with custom RTMP
- Omnichat
- everything in Grow plus webinars for up to 100 registrants
- pre-recorded sessions streamed as live
- recurring webinars
- unlimited sessions
- lead capture
- branded registration forms
- embedding
- reminder and follow-up emails
- unlimited studios and downloads
- up to 10
- 000 webinar registrants
- unlimited video storage and hosting
- private podcast hosting
- custom branding
- async and simultaneous recording
- Producer Mode
Data, GDPR & hosting
A consolidated view of how Riverside handles your data.
GDPR overview
GDPR implementation is concrete and documented rather than merely claimed. The privacy policy, last revised 15 June 2026, names RiversideFM, Inc. as controller, appoints PrivacyTeam Ltd. as external Data Protection Officer reachable at dpo@riverside.com, and designates Alexandre Bouaziz of DEEL IRELAND EOR LIMITED as the Article 27 representative in the European Union, with a separate representative at LETS DEEL LIMITED for the United Kingdom. Transfers rely on certification to the EU-US Data Privacy Framework and its UK and Swiss extensions, on Standard Contractual Clauses approved by the Commission, the ICO and the Swiss FDPIC, and on the adequacy decision covering Israel. A Data Processing Addendum and a named subprocessor list are both published. Access, erasure, portability, objection, restriction and complaint rights are set out and exercised by email.
Who owns the data?
Under the Terms and Conditions, your recordings and files stay yours: the content and any copyright or trademark rights in it remain your property and are used at your sole responsibility. You grant RiversideFM, Inc. a limited, non-exclusive, royalty-free worldwide licence to host, copy, display and reproduce that content, but only as far as delivering the platform requires. Where recordings contain personal data, Riverside acts as processor or service provider under its Data Processing Addendum. Two carve-outs matter: any feedback you send becomes Riverside's exclusive property, and the platform is explicitly not a storage, backup or archiving service, so keeping your own copies is on you.
Reuse rights
You can reuse everything the platform produces without asking Riverside for permission: raw multitrack files, edited exports, transcripts, AI-generated clips, show notes and thumbnails are yours to publish, sell or license, provided you actually hold the rights to what was recorded. The Terms make you warrant that you own or have cleared the content, including consent from every identifiable person appearing in it. On the vendor side the picture is less settled. The FAQ states that Riverside does not own or use your content in any way, while the privacy policy dated 15 June 2026 lists training its own AI and machine-learning models among its processing purposes, on a legitimate-interest basis. Voice and face samples fed to VideoDub are treated more strictly: they serve only the requested feature, subprocessors are contractually barred from training on them, and they are destroyed immediately afterwards.
Data retention & training
Hosting summary
Riverside states that it and its authorised service providers maintain, store and process personal data in jurisdictions including the United States, Canada, the European Union, the United Kingdom, Israel, Mexico, Argentina and the Philippines, among others. Core infrastructure runs on Amazon Web Services and Google Cloud in the United States, with MongoDB as the database provider. The published subprocessor list adds Austria (Auphonic) and Germany (CloudConvert) for media processing, and a series of United States AI vendors including OpenAI, ElevenLabs, HeyGen, Hedra and Pika. Group entities sit in Israel for research and development, Cyprus for payment processing and Canada for sales, with customer support handled from the Philippines. Transfers to Israel rely on the European adequacy decision, transfers to the United States on the EU-US Data Privacy Framework certification, and everything else on Standard Contractual Clauses. No data-residency option letting a customer pin storage to a chosen region is advertised.
Things to keep in mind
Risks and trade-offs to weigh before adopting Riverside.
- The privacy policy permits training Riverside's own AI and ML models on your data under legitimate interest, which flatly contradicts the marketing FAQ and has no documented opt-out; get the position in writing before recording anything confidential
- Voice and face samples supplied to VideoDub may qualify as biometric identifiers under the law that applies to you, which is a heavier category than ordinary recordings
- Regenerating someone's voice and lip-sync after the fact makes it easy to publish words a guest never actually said; treat guest consent as an editorial obligation, not a formality
- Leaning on AI clip selection, show notes and layouts erodes the editorial judgement of a team over time, and the output starts to sound the same as everyone else's
- The platform explicitly states it is not a storage, backup or archiving service, so anything you do not download yourself can be lost
- Monthly download caps and metered AI credits can stall a production mid-cycle, and the true cost only becomes visible once you are already committed
- Disputes fall under the courts of the Tel Aviv District in Israel, with a US arbitration clause and a class-action waiver, which limits your practical recourse
Setup & Integrations
Technical difficulty
Very low for the core workflow. Riverside runs in the browser, so there is nothing to install, and guests join a session from a single link without an account. Optional Mac, iOS, Android and iPad apps exist but are not required, and no particular hardware is imposed, although recording quality follows whatever equipment each participant has. Transcript-based editing is presented as no harder than editing a document. The genuinely technical work is all optional and organisational: brand kits, workspaces, roles, SSO and Producer Mode. Integrating programmatically is the exception, since no public API documentation is available.
Deployment
Apps stores
Integrations
Supported languages
Behind Riverside
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Riverside.
Frequently asked questions
Why does Riverside claim better quality than a video call?
Do my guests need an account or software?
Is the free plan usable in production?
What does the cheapest paid plan cost?
What can the AI actually do?
Are my recordings used to train AI models?
Can I move my edit into Premiere or Final Cut?
What security and compliance does Riverside offer?
Should you pick Riverside?
Riverside has grown well past the remote podcast recorder it started as in 2020. Backed by USD 80 million and used by Spotify, Microsoft, Marvel, TED and The New York Times, it now covers the whole chain from capture to distribution, and it does so around one genuinely good engineering decision: recording each participant locally rather than saving a compressed call. That single choice is what makes the separate 4K tracks, the clean audio and the crash resistance possible, and it remains the clearest reason to pick Riverside over a general video tool. The AI layer on top is unusually broad without being gimmicky, and transcript-based editing puts competent video production within reach of marketers and communications teams who would never open a timeline.
The reservations are mostly commercial. Five overlapping tiers, monthly caps on separate-track downloads, a metered AI credit system and a Business plan priced only on request make it hard to know what you will actually pay before you commit. The free plan is a demonstration rather than a working tier. And developers should note that API access is sold without any public documentation to go with it.
One substantive point deserves attention before signing. The commercial FAQ says Riverside does not use your content, while the privacy policy dated 15 June 2026 lists training its own AI and machine-learning models among its processing purposes, and no opt-out is documented publicly. For a creator that may be acceptable; for an organisation recording confidential interviews or internal communications it is a question to put in writing to the vendor first. Everything else — SOC 2 Type II, ISO 27001, a published DPA, a named subprocessor list and designated Article 27 representatives — is solid.
- Choosing a selection results in a full page refresh.
- Opens in a new window.