📈Last 7 days+84K installs — 40% faster than the week before (+60K) · 6.8K/day since Sep 30
🏆Among Productivity appstop 1% by installs gained this week (614 peers growing)
⭐Rating★ 3.24 → 3.31 since Sep 30
🛠Updatesactively updated · updated 11 days ago · every ~12 days (4 updates tracked)
Estimates · last 30 days ⓘModelled figures, not reported by the store. Use them for scale and trends, not exact numbers.
Downloads ⓘTotal measured from the store install counter over the last 30 days. The split by country is estimated from chart positions in 20 markets with our own model — not figures reported by the store.
144Kmeasured · last 30 days
🇮🇹 Italy35%50K
🇫🇷 France33%48K
Other markets33%
Installs since launch
+244,082 installs
· +373 ratings
since Sep 30
Update history
1.22.0Oct 31, 2026 · seen Nov 5, 2026
- New: support for LFM3 and Bonsai 2 models.<br>- Faster Q2_K and Q3_K models on GPU.<br>- New app icon, with themed icon support on Android 13+.<br>- Fixed: hands-free voice mode could fail without internet for some voices. Voices you've downloaded now work offline.<br>- Updated translations.
1.20.1Oct 26, 2026
New in PocketPal:<br>• Scan a QR code to add a server (remote llama.cpp)<br>• Load and unload llama-server models from your phone (remote llama.cpp)<br>• Web search: added Firecrawl option<br>• Model downloads resume after a dropped connection<br>• Arabic app language<br><br>Fixes:<br>• A crash when a model was unloaded mid-reply<br>• Choosing CPU when loading a model now works<br>• The keyboard c
1.19.0Oct 5, 2026 · seen Oct 16, 2026
Buy Pals directly in the app.
1.18.0Oct 5, 2026
• Faster on Snapdragon NPU: many more models (incl. Q4_K_M/Q6_K) now run on the AI chip, up to 10× faster prompt processing<br>• New language: Spanish<br>• Updated AI engine (latest llama.cpp)<br>• Fixed an NPU crash on some Snapdragon phones<br>• Pal cards display correctly on phones and tablets<br>• Accurate context meter for remote servers (llama.cpp)<br>• Now installable on devices without a c
1.17.3Sep 25, 2026 · seen Sep 30, 2026
• llama.cpp capability discovery: PocketPal now detects a connected server's vision support and context size, and warns when a chat runs out of room.<br>• Internet search in chat: Pals can search the web and read pages. Bring your own key (Brave, Tavily, Exa).<br>• Speculative decoding (experimental): pair a small draft model for faster replies.<br>• Pin chats, export chats as Markdown.<br>• P
PocketPal AI is a productivity app by LLM Ventures with 1,000,000+ downloads (about 3.1M estimated). +84K installs in the last 7 days. Rated 3.3 from 3.8K ratings. Released Sep 11, 2024. Install growth, rank history, SDKs and ad networks.
From the store listing: “Your AI. No one else's.
PocketPal runs large language models directly on your phone. No servers. No internet. No data leaves your device. Ever.
This is AI you own — not AI you rent.
WHAT YOUR PHONE NEEDS
Modern phones are powerful enough to run real AI. PocketPal uses that power.
You'll need 6GB+ RAM for smaller…”
