Off Grid AI
Off Grid AIOff Grid AI is an on-device assistant for iPhone and Mac. It runs GGUF language models locally with vision input, Whisper speech recognition and image generation, and adds retrieval over local files and MCP tool access. A paid Pro tier adds on-device text-to-speech, custom personas, approval-gated actions and cross-device sync.
Verified 2026-08-31 via the repository LICENSE and README.
Openness
4 high confidence- license
- MIT(OSI)
- source
- public(off-grid-ai/OGAM)
- core-gated
- gated(the README sells an Off Grid AI Pro tier at 69 dollars for life or 49 dollars a year, and lists the capabilities it adds over the published app - on-device text-to-speech, custom personas, approval-gated actions through Calendar, email and MCP servers, and cross-device sync)
MIT, an OSI licence, with the full source public and the shipped code in this repository. The LICENSE body was read in full: the stock MIT text, with no appended condition. The gating sits in the product's pricing rather than in its licence. The readme sells an off grid ai pro tier at 69 dollars for life or 49 dollars a year, and lists the capabilities it adds over the published app - on-device text-to-speech, custom personas, approval-gated actions through calendar, email and mcp servers, and cross-device sync.
- https://raw.githubusercontent.com/off-grid-ai/OGAM/main/LICENSE recorded 2026-08-31
LICENSE body, read in full: the stock MIT licence text carrying a 2026 copyright line for Mohammed Ali Chherawalla, with no appended term or additional condition.
- https://raw.githubusercontent.com/off-grid-ai/OGAM/main/README.md recorded 2026-08-31
README, read for the distribution and gating question. The 'Off Grid AI Pro' section prices a tier at 69 dollars for life or 49 dollars a year and its 'What Pro adds' list names capabilities the published app does not have: on-device Kokoro text-to-speech, custom personas with their own system prompt, voice and persistent memory, drafting actions against Calendar, email and MCP servers for approval, and phone-to-Mac sync. Those are builds of this same product rather than a service sold beside it, so the core reads as gated.
Adoption
2 low confidenceBanded on stars, because the product ships through an app store rather than any registry the routing table bridges. 3,030 stargazers places it at level 2, and the instrument caps the band at 3. Stars measure attention rather than use, so confidence is low.
- https://api.github.com/repos/off-grid-ai/OGAM recorded 2026-08-31
Repo metadata for off-grid-ai/OGAM: stargazers_count 3030, archived false, pushed_at 2026-08-30.
Capability
3 medium confidenceScored 3 in the local-application class alongside jan. Its modality range - text, vision, speech in, image generation, all on device - is wider than most of that band and is what holds it at the top of it, but it targets one person's phone and laptop rather than providing the multi-backend, document and agent platform surface that places open-webui and anythingllm at 4.
- https://raw.githubusercontent.com/off-grid-ai/OGAM/main/README.md recorded 2026-08-31
README documents on-device GGUF language models, vision, Whisper speech recognition and image generation for iPhone and Mac, retrieval over local content, MCP support, and a Performance section reporting on-device figures.
Verified 2026-08-31