February 14, 2026
Off-grid app, on-fire comments
Show HN: Off Grid – Run AI text, image gen, vision offline on your phone
Offline AI on your phone sparks cheers, broken‑link rage, and F‑Droid purists waiting
TLDR: Off Grid promises full offline AI—chat, images, vision, and voice—running on your phone without the cloud. The crowd cheered privacy and design, but clashed over a broken download link, Android users waiting for F-Droid, and iOS sideload headaches, turning a neat launch into a lively URL and app-store showdown.
“Your AI, your data” went full off‑grid today with a pocket‑sized suite that chats, sees, listens, and paints pictures entirely on your phone—no cloud, no subscription. Fans swooned over the ethos and clean UI: one commenter praised it as “well thought out” and loved the privacy-first vibe. Performance boasts are bold—text replies in a blink, images in seconds, and voice-to-text on the fly—making phones feel like mini studios. But the real show? The crowd drama.
Cue Dash‑Gate: the GitHub Releases link initially broke because of a missing dash, sending download-hungry users into a frenzy while sleuths posted fixes like heroes in a URL thriller. The Android crowd split between “install the APK now” and the F‑Droid faithful waiting for a build from the open-source app store. On iOS, confusion ruled: how do you even sideload (install outside the App Store)? That sparked jokes about an “off‑grid app stuck behind Apple’s fence,” and quips like “Swiss Army Knife, but the corkscrew is missing.” The mood: big excitement for offline AI superpowers, spiked with broken-link chaos and store‑politics debates. It’s privacy poetry—plus a dash that launched a thousand comments.
Key Points
- •Off Grid is an on-device, offline AI app for mobile that provides text generation, image generation, vision, voice transcription, and document analysis.
- •Text generation supports Qwen 3, Llama 3.2, Gemma 3, Phi-4, and GGUF models with streaming responses and ~15–30 tok/s on flagship devices.
- •Image generation uses Stable Diffusion with NPU acceleration on Snapdragon and Core ML on iOS, typically producing images in 5–10 seconds on flagship hardware.
- •Vision AI employs SmolVLM, Qwen3-VL, and Gemma 3n for scene and document analysis; Whisper handles real-time on-device speech-to-text.
- •The app is open-source with an Android APK and build-from-source instructions for Android/iOS; performance was tested on Snapdragon 8 Gen 2/3 and Apple A17 Pro, and detailed documentation is provided.