On this page
You can run LFM3 on your iPhone with Off Grid AI Mobile (OGAM). Download one model of about 1 GB once, and you get a fast chat model, or one that reads your photos, running on your own iPhone with no internet and no uploads.
It’s free, and it fits iPhones with 4 GB of RAM, so an iPhone 12 or 13 runs it as well as a new Pro.
App Store · Download Off Grid AI
What LFM3 is
LFM3 is Liquid AI’s newest small model family, released in October 2026. It mixes short convolution layers with attention layers, a design Liquid built to run on phones and laptops. Liquid’s model card says the 1.7B model takes about 1 GB of memory at 4-bit, which is the build OGAM downloads.
OGAM has two LFM3 models in its catalog:
| Model | What it reads | Download | RAM listed | Fit on iPhone |
|---|---|---|---|---|
| LFM3 1.7B | Text | 1.1 GB | 4 GB or more | Runs well on 4 GB and up |
| LFM3 VL 2B | Text and photos | 1.5 GB with its vision file | 4 GB or more | Runs well on 4 GB and up |
Both download at Q4_K_M. The VL model comes as two files, the model and a vision file (an “mmproj” file). OGAM fetches them together and pairs them when the model loads.
On iPhone, OGAM marks a download under 60% of RAM as running well. LFM3 VL 2B at about 1.5 GB sits well under that even on a 4 GB iPhone, so it leaves room for a long chat and your other apps.
What LFM3 1.7B is good for
OGAM’s catalog describes LFM3 1.7B as “Fast agent and tool use with reasoning; text only”. Liquid trained it for tool calling and structured answers, and it understands 16 languages, including Hindi, Arabic, Japanese, Spanish and Russian. Try:
- “Turn these notes into a packing list, grouped by bag.” Paste messy text and get it back in order.
- “Reply to this message politely and say I can do Thursday instead.” A short draft you can paste into Messages.
- “Put these expenses in a table with a total.” Structured output is what it was trained for.
- “Rewrite this in simple Hindi.” Hindi is one of its 16 languages.
Liquid says the 1.7B model isn’t meant for agentic coding or hard multi-step reasoning. If your iPhone has 8 GB or more and you want a stronger all-rounder, the Qwen 3.5 iPhone guide covers the next size up. LFM3 is the pick for quick answers on an older iPhone.
What LFM3 VL 2B does with photos
LFM3 VL 2B is the LFM3 language model with a 164M-parameter vision encoder added. The catalog line is “Fast reasoning with vision, reads documents and charts”, and Liquid lists reading documents, charts and app screens as what it’s built for. Take a photo or pick a screenshot, then ask:
- A printed bill or receipt: “What’s the total, and how much of it is tax?”
- A screenshot of a chart: “Which month had the highest number?”
- A page of a letter or form: “What’s the deadline in this letter?”
- A photo of a shelf or a desk: “Where is the blue folder?” Liquid trained it to find objects and say where they sit.
The photo stays on your iPhone. A cloud assistant has to upload your bill or your letter to answer, and OGAM doesn’t.
Set it up
- Install OGAM from the App Store. You need version 0.0.122 or later.
- Tap Models in the tab bar and stay on the Text tab.
- Type LFM3 in Search models.
- Tap LFM3-1.7B-GGUF for chat, or LFM3-VL-2B-GGUF if you want to ask about photos. Both are Liquid AI’s own uploads. You can also open the filters, tap Org and pick Liquid AI.
- Under Sizes, take the Recommended file and download it. Use Wi-Fi, because it’s 1.1 to 1.5 GB.
- When it finishes, tap Chat now.
To ask about a photo with LFM3 VL 2B, tap + in the message box, choose Photo, then Camera or Photo Library. Add your question and send. If the app says “Vision Not Supported”, LFM3 1.7B is still the active model, so switch to LFM3 VL 2B.
Once it’s downloaded, turn on airplane mode and send another message. It still answers, because the model runs on your iPhone.
If something goes wrong
| Symptom | Check | Fix |
|---|---|---|
| No LFM3 results | Your app version | Update OGAM to 0.0.122 or later from the App Store. LFM3 needs the newer llama.cpp engine that version ships |
| “Vision Not Supported” when you add a photo | Which model is active | Switch to LFM3 VL 2B |
| “Vision File Missing” | The model is there, its vision file isn’t | Open your downloads from that message, tap the wrench next to LFM3 VL 2B, then reload it |
| Long, wandering answers on a hard problem | LFM3 is a small model | Ask shorter questions, or use Qwen 3.5 if your iPhone has the RAM |
FAQ
Can LFM3 run on an iPhone? Yes. OGAM 0.0.122, the current App Store version, runs LFM3 1.7B and LFM3 VL 2B. Both are listed for 4 GB of RAM, so they fit an iPhone 12 or newer, and both run with no internet.
How big is the LFM3 download on iPhone? LFM3 1.7B is about 1.1 GB at Q4_K_M. LFM3 VL 2B is the same size plus a vision file of about 0.38 GB, so about 1.5 GB in all. The app fetches both files together.
What is the difference between LFM3 1.7B and LFM3 VL 2B? LFM3 1.7B reads text only. LFM3 VL 2B is the same language model with a 164M-parameter vision encoder added, so you can attach a photo or screenshot and ask about it.
Is LFM3 free on iPhone? Yes. Chat and photo questions are part of the free Off Grid AI app, with no account and no subscription.
Does LFM3 work offline on iPhone? Yes, after the first download. The model runs in your iPhone’s memory, so you can turn on airplane mode and keep chatting. Your messages and photos aren’t uploaded.
Related guides
- How to Run LFM3 on Your Android Phone in 2026 (1.7B Chat and 2B Vision, Offline)
- How to Run Qwen 3.5 on Your iPhone in 2026 (Completely Offline, No Subscription)
- How to Find AI Models Your iPhone Can Actually Run in 2026 (No Failed Downloads)
- How to Ask AI About a Screenshot on iPhone in 2026 Without Internet
Get Off Grid AI on the App Store and run LFM3 on your own iPhone today.