OneLLM
OneLLM OfficeConnectSupport
English ▾
English简体中文EspañolDeutschFrançaisالعربية
Get app ↗

ONELLM OFFICE + ONELLM MOBILE

Your AI.
Ready for real work.

Create reports, presentations and spreadsheets. Choose local or online AI, then keep working from your phone.

  • Conversations saved on your device
  • Local & online models
  • Files you can review and use
OneLLM Office on MacOneLLM on iPhone: local AI, online models and computer connections

ONE WORKSPACE. YOUR CHOICE.

On your device.
Connected to your world.

Keep a conversation local, choose your own cloud provider, or continue an agent running on your computer.

ONELLM OFFICE

From scattered notes
to a useful next step.

Explore the new OneLLM Office workspace. Clear plans, useful reports and polished first drafts, starting with a familiar task.

Explore OneLLM Office →
OneLLM Office on Mac

LOCAL AI

35B

Room for bigger thinking.

Run a supported 35-billion-parameter model on your device with expert streaming. Download once, then chat offline.

  • Private on-device inference
  • Smaller text and vision models, too
  • A shared chat and model library

The optional 35B MoE package has about 3B active parameters and requires about 16.6 GB of storage. 12 GB RAM is recommended. Performance varies by device and model.

Explore local AI ↗
OneLLM model library with the optional 35B package

YOUR COMPUTER. WITH YOU.

Your agents,
within reach.

Start in OneLLM Office. Continue from OneLLM on iPhone or iPad. Connect to OpenCode and DS Harness, too.

OneLLM Office

Continue tasks, choose AI, review approvals and collect finished files from your phone.

OneLLM mobile with the OneLLM Office connection

OpenCode

Continue sessions, choose models, browse files, review changes and respond to your agent.

An actual OpenCode conversation in OneLLM

DS Harness

Bring conversations, workspaces, photos, files and agent controls into a compact mobile view.

An actual DS Harness conversation in OneLLM
Get OneLLM Connect →Keep your computer awake. Connect over Wi-Fi or Tailscale.

LESS FRICTION. MORE FLOW.

The details make it easy.

See the conversation.

Images appear right in chat. Preview an attachment before sending it, or remove it with a tap.

Say it naturally.

English voice input with offline transcription. Speak, check your words and send.

Choose the right help.

Use local AI for private work, or choose from leading online models. Your task stays in one workspace.

Offline AI on iPhone and iPad.

Download a supported model in OneLLM while online, then choose that local model to chat without an internet connection. Online models and computer connections still need a network.

How do I start an offline conversation?

Open the model library, download a model that fits your device, and select it before starting a chat. Try a short message in airplane mode to check your setup.

Which model should I choose?

Start with a smaller model and check available storage and memory. The optional 35B MoE package needs about 16.6 GB of storage; 12 GB RAM is recommended. Speed varies by model and device.

Can I also use OpenCode from my phone?

Yes. OneLLM can connect to OpenCode running on your computer. This is a separate connection: the computer must remain on and reachable.

Set up OpenCode on iPhone →

PRIVATE BY CHOICE

Make space
for your next idea.

Get OneLLM ↗