Free alternatives · Video, slides and media
ShortsFromVideo
Turn one long video into ranked, reframed, captioned short clips, plus a blog post, LinkedIn post and X thread.
A free alternative to OpusClip, built by Surge Software. Free. You pay only what third-party services cost us, at cost.
Live at shortsfromvideo.com
The problem
What it solves for a business
For: Podcasters, YouTubers, educators and teams who record long videos (podcasts, lectures, interviews, live streams) and need short clips from them.
- Finding the best moments in an hour-long recording and cutting them by hand takes hours.
- Clipping tools burn credits on re-runs and bad clips, and monthly credits expire.
- AI clips often cut before the punchline, and the crop jumps to the wrong person when several people talk.
How it works
ShortsFromVideo, step by step
- 01
Add a long video
Upload a podcast, lecture, interview or stream recording.
- 02
Transcribe
Speech is turned into text with word timings on our own servers.
- 03
Find the moments
Sentence-aligned windows are scored by code, then re-ranked by our own AI model, and each clip gets a score with a plain reason.
- 04
Render clips
Each clip is reframed on the active speaker, gets animated captions and a hook banner, and has filler words and silences cut.
- 05
Edit and reuse
Fix clips in the clip editor and get a blog post, LinkedIn post and X thread from the same recording.
Comparison
ShortsFromVideo vs OpusClip
| ShortsFromVideo | OpusClip | |
|---|---|---|
| Price | Free. You pay only what third-party services cost us, at cost. | Starter $15/month, Pro $29/month (checked 24 Oct 2026) |
| Cost per source minute | $0 for our own processing (600 source min/month fair use) | $0.097 to $0.100 |
| Free plan | Every feature, no watermark | 60 min, watermark, files expire after 3 days |
| Re-edits and re-renders | Free | Cost credits; users report credit burn on re-runs |
| Credits | Prepaid at-cost credit stays in your account | Monthly credits expire after 60 days; no plain top-up (third-party source) |
| Clip cut points | Sentence-aligned windows | Reviews report cuts before the punchline |
| Speaker tracking | Switches only after 1 s and a clear lead, then cuts | Reviews report the crop jumping to the wrong person with 3+ speakers |
| Podcast bundle (blog, LinkedIn, X thread) | Included | Not offered |
OpusClip details are taken from its public website and may have changed since (Starter $15/month (150 min), Pro $29/month (300 min); free plan 60 min with watermark (checked 24 Oct 2026)).
Why switch
Advantages over OpusClip
No credit maths
Our own processing is free up to 600 source minutes a month. Re-edits and re-renders cost nothing.
Clips that finish the thought
Candidate clips start and end on sentence boundaries, and each one shows why it scored.
Steadier speaker crop
The tracker holds on a speaker for at least a second and needs a clear lead before it switches.
More from one recording
One podcast gives clips plus a blog post, a LinkedIn post and an X thread.
Features
What it does today
- AI highlight detection on sentence boundaries, so clips end on a full thought.
- Virality score 0 to 100 with four sub-scores (hook, flow, value, trend) and a plain-language reason.
- Auto-reframe to 9:16 and 1:1 with active-speaker tracking that waits before switching, plus a split layout for two speakers.
- Animated captions (karaoke, pop, bounce, boxed, minimal) with emoji, keyword colour and safe zones per platform.
- Filler-word and silence removal with captions retimed, and a hook banner in the first seconds.
- Podcast bundle: an SEO blog post, a LinkedIn post and a 6-post X thread from the same recording.
- Clip editor, analytics page and cost per clip shown in rupees and dollars.
- Team workspaces with roles and a REST API with hashed API keys.
Use cases
How teams use ShortsFromVideo
Weekly podcast
A two-host podcast becomes a set of 9:16 clips with captions and a blog post the same day.
Lecture highlights
A teacher turns a recorded class into short clips that each end on a complete point.
Founder interviews on LinkedIn
A startup team cuts 1:1 clips from an interview and gets a ready LinkedIn post to go with them.
Under the hood
Where AI is used, and where it is not
The model
Speech-to-text runs on whisper.cpp on our own servers. Clip re-ranking and the podcast bundle use Qwen (qwen3.8-27b) on the Surge AI server first, with Claude Sonnet as fallback only when Qwen fails. About 2 model calls per project.
Rules and code
Code builds sentence-aligned candidate windows, scores 9 features, and checks model output with a schema and range guard; an offline ranker is used if the model fails. Reframe, captions, filler cuts and renders are deterministic code (ffmpeg).
People in the loop
The user reviews the ranked clips, edits them in the clip editor and decides what to download or schedule.
Book a working session
A working session with our architects on one of your use cases. You bring a process and sample documents; we leave you with a scoped plan: model, hosting, integrations, timeline and cost range.