GPT-Video vs Riverside: clips from any file, counted in minutes of source
Riverside is a recording studio that added clips and sells hours of recording, while GPT-Video is a clip editor that sells minutes of a file you bring.
GPT-Video starts from any video file you already have. There is no studio to book and no recording hours to use up, and a credit is one minute of the file you upload. Starter is $19 a month for 250 credits. Riverside records the interview itself, has a free plan and, on Grow and above, posts for you through its MCP server.
The short answer
Riverside and GPT-Video meet at one step, which is turning a long conversation into short vertical clips. They start from different places. Riverside is where you record the conversation, and the clips are one feature of that studio. GPT-Video does not record anything. You bring a finished file from wherever it was made, and the whole product is about what to cut from it.
If your episodes are already on a drive, that difference decides most of the choice. You pay for minutes of the file, you do not need a studio session, and nothing on the plan is sized around hours you never spent recording. If you want the recording tool too, Riverside has it and we do not.
Where GPT-Video is stronger
It starts from any file. Upload a video made in any recorder, a phone, a conference call or an old archive, and GPT-Video works on it. There is no account of recording hours to run down first. Riverside's Magic Clips page says clips come from a recording or an uploaded video, so uploads are possible there too. What its pricing page does not say is whether an uploaded file uses up the monthly recording hours. We could not confirm it, so check before you plan around it.
The meter counts the file. Each credit pays for a minute of your upload, and Standard is the base quality. A Draft pass halves the cost and a Maximum pass multiplies it by 2.5. Add up the length of your episodes and you know the month. Riverside's plans are built on hours of recording, hours of separate-track downloads and a pool of AI credits. Its pricing page gives no clip count or minute limit for Magic Clips, so the clip allowance cannot be priced from the page.
The picture follows the speaker. GPT-Video reframes on the speaker's face for the 9:16 crop. Riverside's Magic Clips page lists 16:9, 1:1 and 9:16 outputs and reframing by speaker or topic. Face tracking is not described on the pages we read. The details are on the auto reframe page.
Each pick comes with a reason. The clip card lists the hook it begins with and the spoken line behind its score, so disagreeing with a pick takes a few seconds.
You edit by describing. Tell Editor Assistant to move the captions up or trim the first pause, and the clip on screen changes. You can undo each step. More on vibe editing, and a walk-through for podcast clipping.
Side by side
| Feature | GPT-Video | Riverside |
|---|---|---|
| Main job | Clips from a video file you already have | Recording studio with an editor and Magic Clips |
| Records interviews | No | Yes, separate audio and video tracks, up to 4K |
| Entry paid plan | Starter, $19 a month, 250 credits | Pro, $29 a month, 5 hours of recording |
| Plan above that | Creator, $49 a month, 800 credits | Grow, $39 a month, 15 hours of recording |
| What the plan counts | Minutes of source, one pool | Hours of recording, hours of track downloads, AI credits |
| Free plan | No permanent plan; free access for new accounts: 50 credits for 14 days, no card | 2 hours of multi-track recording, one time, up to 720p |
| Face reframe | Centers on the speaker's face | Reframes by speaker or topic; face tracking not described |
| Why a clip was picked | Hook and line shown with each clip | Not confirmed on their public pages |
| Chat editing | Editor Assistant, each change undoable | Text-based editing; chat editing not confirmed |
| Publishing and scheduling | None | Yes, through its MCP server on Grow and above |
| MCP | Eight tools to find and export clips | Yes, for Claude, ChatGPT and Cursor |
| Public REST API | No | Listed for the custom Business plan only |
Where Riverside is stronger
- Recording. It captures the interview itself, with separate tracks for each guest and video up to 4K. GPT-Video only works on video you already have.
- A free plan. You get 2 hours of multi-track recording once, with Magic Clips included, at up to 720p. GPT-Video has no free plan. A new GPT-Video account only gets a fortnight of free access with 50 credits.
- Posting. Through the MCP server an assistant can queue posts for YouTube, TikTok, Instagram, Facebook and X, and draft titles, thumbnails and show notes. GPT-Video stops at the exported file.
- Voice repair. VideoDub regenerates spoken words in the same voice, and layouts switch to whoever is talking.
- An API. It exists on the custom Business plan. GPT-Video has none.
What a real month costs
Take a podcaster who has 4 hours of finished episodes a month. That is 240 minutes, so 240 credits at Standard. Starter holds 250 and costs $19 a month. Annual billing pays for 10 months and gives 12, which makes it $190 a year.
Riverside Pro is $29 a month, or $24 a month on annual billing, which is $288 a year. Against Starter that is $10 more each month and $98 more over a year. Pro includes 5 hours of recording, which is 300 minutes, so $29 for 300 minutes is about $0.097 a minute of allowance. A Starter minute is $0.076. Those two minutes are not the same thing, because Riverside's hours are recording hours. Its price also covers a studio that you may or may not use.
Now a heavier month, with 12 hours of episodes, or 720 minutes. Adding one top-up to Starter gives 750 credits for $44 a month. Creator has 800 credits for $49 a month and works out to about $0.061 a minute. On the 10-for-12 terms Creator is $490 a year. Riverside Grow is $39 a month for 15 hours of recording, or $34 on annual billing, which is $408 a year. Here Riverside is $5 to $10 a month lower and $82 a year lower than Creator, and it includes the studio. If you would record there anyway, Grow is the cheaper path. If you would not, you pay for the studio and do not use it.
What you give up by moving
You give up the recording room, the free plan, the voice repair and the posting through MCP. You post the exported files yourself, or hand them to the scheduler you already use.
You get a plan that is sized by the length of your files, a reframe on the speaker's face, a reason next to each clip and a chat that edits the clip in front of you. If your episodes are recorded already, sign up and confirm your email. Places for free access are limited, and each one carries 50 credits and a 14-day limit, with no card taken. Upload an episode of up to about 50 minutes, read the picks, and judge the clips on your own material. Every plan and its credits are listed on the pricing page, and the MCP connector is described in the MCP docs.
The verdict
Which one to pay for
Choose Riverside if
- You record remote interviews and want the studio to capture separate audio and video tracks, up to 4K.
- You want a free plan that records as well as clips. It gives 2 hours of multi-track recording, one time, at up to 720p.
- You want an assistant to schedule and post for you. Its MCP server does that on Grow and above.
- You edit by text, fix spoken words with AI voice regeneration, or want layouts that follow the active speaker.
Choose GPT-Video if
- Your episodes are already recorded, in any tool, and you only need the clips.
- You would rather pay for minutes of source than for hours of recording you did not make in a studio.
- You want each clip centered on the speaker's face, with the reason for the pick written next to it.
- You like to adjust a clip by telling Editor Assistant what to change.
Prices and features on this page were read from each product's own public pages on 4 October 2026. Prices move; if something is out of date, it is a mistake and not a tactic.
Questions
Is GPT-Video a good Riverside alternative?
It is a good alternative if your recording happens somewhere else and you want the long file turned into vertical clips. It is the wrong one if you need a place to record the interview, because GPT-Video only works on video you already have.
Can I use GPT-Video without recording in a studio?
Yes, that is how it works. You upload any finished video file and it finds the clips. There is no recording room, no booking and no guest link. You cannot paste a link, so the file itself goes in.
How does GPT-Video charge compared with Riverside?
One credit is one minute of the uploaded file at Standard quality, from a monthly pool. Riverside sells hours of recording per month, plus hours of separate-track downloads and a separate pool of AI credits for features like translation.
Does GPT-Video post my clips like Riverside's MCP server can?
No. GPT-Video ends at the exported file and never connects to your social accounts. It has an MCP connector for Claude, ChatGPT and Cursor, with eight tools to find and export clips, and you post the files yourself.
Does GPT-Video have a free plan or trial?
It has free access, which is not a plan. A new account that confirms its email gets 50 credits for 14 days, with no card, while places last. Then Starter at $19 a month is the first paid plan. Riverside lists a free plan with 2 hours of multi-track recording, given one time.
Related
Try it on your own footage
Every plan comes with the whole editor. The only difference is how much video you run through it.
Plans from $19 a month. Cancel any day, keep the month.