SAASPOCALYPSEverdict #LOOM-8488
scanned 2026.04.27 · 15:11
subject of investigation
loom.com
▸ async screen recording & video messaging
verdict: MONTH
buildability score
38
/100
tier · month
the blunt take
“The recording button is a weekend. The latency is a month. The AI bug reports are someone else's month. You're not cloning Loom — you're cloning the "it just works" part, and that's the whole ballgame.”
Screen capture APIs exist, upload pipelines exist, transcription is a commodity — but stitching them into something that feels instant across Chrome extensions, desktop apps, and mobile, with real-time processing and viewer analytics, is a genuine multi-week engineering slog with several sharp edges.
cost breakdown.
their price ←→ your price
what they charge●
Starter plan
$15
/ user/mo
※ Free tier available; Business plan ~$15/user/mo billed annually
annual:$180
what it costs you✦
01 · Cloudflare R2 (video storage, ~500GB)$8.00
02 · Cloudflare Stream or Mux (video encoding + delivery)$55.00
03 · OpenAI Whisper API (transcription, ~10hrs/mo)$12.00
04 · Vercel Pro (web app + API routes)$20.00
05 · Supabase Pro (auth, metadata, viewer events)$25.00
06 · Resend (share notifications)$0.00
07 · Domain + DNS$2.00
08 · OpenAI GPT-4o (AI summaries + bug report gen)$5.00
TOTAL / mo127
▸ break-even:4 users on Business plan (~$15/user/mo) covers ~$60/mo infra. Real break-even is around 10 seats.
or, you know, use one of these.
if building feels spicy
option A
OBS + Notion (self-hosted workflow)
Free. Record with OBS, upload to Notion or Google Drive, share the link. Covers 80% of use cases for $0.
option B
Cap (open source Loom clone)
It literally exists. Actively developed. Self-host or use their free tier. Don't reinvent this wheel.
option C
Screenpal (formerly Screencast-O-Matic)
Cheaper paid tier, solid enough for async comms, no build required.
what'll actually be hard.
est. total: 8 weeks
▸ 1 week Chrome extension + desktop recording · 2 weeks upload pipeline + video processing · 1 week sharing/embeds + viewer · 2 weeks transcription + captions + AI summary · 1 week video editor UI · 1 week auth, analytics, polish
easy
medium
hard
nightmare
01
easy
Video upload + storage pipeline
Presigned S3/R2 URLs + multipart upload. Plenty of tutorials. Done in an afternoon.
02
easy
Shareable video page with viewer
HLS player (Video.js or Mux Player), public URL, basic auth guard. Straightforward CRUD.
03
medium
Chrome extension for screen capture
getDisplayMedia() API works, but manifest v3 + background workers + reliable stop/save flow has enough edge cases to ruin a weekend.
04
medium
Transcription + 50-language captions
Whisper handles this well, but syncing captions to video timestamps and rendering them on the player is fiddlier than it looks.
05
hard
Video processing pipeline (trim, stitch, overlays)
FFmpeg on a worker queue (BullMQ + Lambda or Modal). Trim is easy. Stitch + text overlays + custom backgrounds = hours of FFmpeg flag archaeology.
06
nightmare
Sub-second upload-to-playback latency
Loom's killer feature is that the link works before the upload finishes. Implementing progressive HLS chunking mid-upload is genuinely hard and requires a real media pipeline, not just 'upload then process'.
recommended stack
Next.js 15 + Vercel (web + API)Cloudflare R2 + Mux (storage + streaming)FFmpeg on Modal or AWS Lambda (processing)OpenAI Whisper + GPT-4o (transcription + AI features)Supabase (auth + Postgres + realtime viewer events)
ready to build?
We'll email you the build guide. Cancel some plans.
▸ generated with love, by a heartless robotverdict v2.1 · saaspocalypse.dev