I had 2,847 bookmarks in Pocket when it died.
I'd read maybe 94 of them. And I'm the guy who built a read-later app.
When Mozilla killed Pocket on July 8, 2025, I wasn't sad — I was relieved. My unread pile finally had an excuse to disappear. But then I realized: if I just move 2,847 links to another app, I'll end up in exactly the same place.
That realization is why I built Burn 451. But this post isn't a sales pitch — it's the honest guide I wish I had when I was looking for alternatives. I've tested all of them. Some are better than Burn for certain workflows. I'll tell you which.
No affiliate links. No sponsored placements.
What happened to Pocket?
Pocket was shut down on July 8, 2025 after Mozilla cut investment over two years. 20+ million users got an HTML export file and a "good luck."
Here's the timeline:
- 2007 — Read It Later (later Pocket) launches as a bookmarklet
- 2017 — Mozilla acquires Pocket, integrates into Firefox
- 2023 Q4 — Mozilla layoffs hit the Pocket team hard
- 2024 Q1 — Pocket Premium killed, free-only
- 2025 March — Mozilla announces sunset
- 2025 July 8 — Pocket dies
- 2025 October — Export tool goes offline. Data gone forever.
If you haven't exported yet and it's before October 2025: do it now at getpocket.com/export.
So what should I use instead?
Short answer: it depends on what you actually do with saved articles. Most people save and never read — which means the tool doesn't matter. What matters is whether the tool changes that behavior.
Here's my honest ranking after testing everything:
| App | Price | AI | CLI/MCP | Best For |
|---|---|---|---|---|
| Burn 451 | Free / $4.99/mo | Full AI digest + MCP | Yes | People who hoard and never read |
| Raindrop.io | Free / $3/mo | Search only (Pro) | API only | Pure bookmark organizing |
| Readwise Reader | $8/mo | Ghostreader highlights | API | Serious highlighters |
| Instapaper | Free / $3/mo | None | API | Minimalists |
| Matter | Free / $8/mo | Basic | No | Apple ecosystem |
| Wallabag | Free (self-host) | No | API | Privacy-first |
| Karakeep | Free (self-host) | AI tagging | API | Self-hosters + AI |
| GoodLinks | $5 once | No | No | Apple-only, no subscription |
A few notes:
Omnivore was another promising open-source option. ElevenLabs acquired it in late 2024 and killed it almost immediately. "Open source" doesn't guarantee longevity when the team pivots.
Pocket's real gap wasn't features. It was the largest free read-later app with deep browser integration. Nothing fills that exact hole, but honestly? The alternatives are all more capable.
For developers: CLI + MCP is the real story
If you write code and use AI tools, the MCP integration is the thing that actually matters. I'm biased here — I built it — but hear me out.
Your saved articles can become context for Claude, Cursor, or any MCP-compatible AI. You don't copy-paste links. You ask "what did I save about WebSocket performance?" and get answers from your own reading history.
- Burn 451: Full CLI + 22-tool MCP server + REST API
- Wallabag: Self-hosted, full API, no AI (build your own)
- Raindrop.io: Clean API, no CLI, no MCP
If you live in the terminal and work with AI, Burn's ecosystem is genuinely different. If you want total infra control, Wallabag.
The free tier showdown
Raindrop has the most generous free tier for organizing. Burn has the most generous free tier for actually reading.
| Burn 451 Free | Raindrop Free | Wallabag | |
|---|---|---|---|
| Unlimited saves | ✓ | ✓ | ✓ |
| AI features | ✓ | ✗ | ✗ |
| Full-text search | ✓ | ✗ (Pro) | ✓ |
| CLI/MCP | ✓ | ✗ | ✗ |
| Mobile | iOS | iOS + Android | Both |
| Setup | Zero | Zero | Need a server |
Burn's free tier has a catch: the 24-hour burn timer. Articles expire if you don't act on them. This is by design — it's a digestion system, not storage. If you want 10,000 articles sitting quietly forever, use Raindrop or Pinboard.
How to export your Pocket data
Go to getpocket.com/export → download HTML → import to your new app. Deadline: October 2025.
For Burn 451:
npm install -g pocket-to-burn
pocket-to-burn import --file pocket_export.html
Tags, timestamps, and read/unread status are preserved. Old articles go straight to Spark (not the 24-hour inbox).
Most other apps (Raindrop, Readwise, Instapaper, Wallabag) also accept Pocket HTML import directly through their settings.
What makes Burn 451 different
Most read-later apps compete on the same axis: better storage, better folders, better reading font. Burn rejects the premise. The problem isn't saving — it's that you save too much and never go back.
- 24-hour timer: Every article gets a countdown. Read it, vault it, or let it burn.
- AI digest: Don't read 15 articles — read a 2-minute synthesis of what matters.
- MCP server: Your reading history becomes context for AI tools.
The metaphor: information comes in, gets processed, nutrients absorbed, waste eliminated. Hoarding is constipation.
Is it for everyone? No. If you want a quiet archive, use Raindrop or Pinboard. Burn is for people who are tired of lying to themselves about "reading it later."
Where Burn falls short (yes, I'm listing my own weaknesses)
- No Android app — iOS and Web only as of April 2026
- Opinionated — The timer stresses some people out
- Young product — Launched 2025, less polished than decade-old competitors
- Not for archivists — If you want 10K bookmarks searchable in 3 years, that's not us
The bottom line
Pocket's death is a forcing function. Don't just migrate your pile to a new app. Ask yourself: do I want a storage system or a processing system?
- Storage: Raindrop.io (free) or Pinboard ($22/year)
- Premium reading: Readwise Reader ($8/mo)
- Self-hosted: Wallabag or Karakeep (free)
- Actually read what you save: Burn 451 (free)
94 out of 2,847. If you see yourself in that number, maybe give it a shot.
