oneira
OverviewFeaturesCompareSpecsPreorder a Founding Unit · $80
Oneira guides/Agent controller

Agent guide / Connected beta, intent before autonomy

Optional connected workflows—with bigger work ahead.

See how optional compatible connected tools and services fit beside Oneira's local experience, with clear permission and capability boundaries.

Preorder a Founding Unit · $80 →See current features →

Product status

What is available now?

  • Working today

    Working today means the expressive face has been demonstrated on a real Oneira device.

  • Connected beta

    Connected beta features require the Oneira app, internet access, a compatible service, and permissions you approve. They can change and may not work with every service.

What you need

Connected beta features need the Oneira app, internet access, a supported service, and permissions you approve. A separate account or fee may also be required.

The short answer

Oneira supports bounded preview tasks, while optional compatible connected tools can extend the experience. Browsing, coding, and workspace changes are not current included capabilities.

Real Oneira product resting in an open palm with its exterior visible
Real development hardware / control layers
PreviewBounded tasks
SetupOwner configured in the app
Connected toolsOptional / owner configured
Research + codeIn development

Oneira Dream Big+

A local AI powerhouse in a star.

Small enough to look playful. Built with a dual AI chip, OLED screen, and 1K camera, plus optional connected experiences you control.

Each $80 Founding Unit includes one Oneira device with a dual AI chip, OLED screen, and 1K camera, plus the selected color and included eye set. The 1K camera is included; its final production module and other non-material production details may change before fulfillment. Optional connected services, provider accounts, usage, and third-party fees are separate.

On-device

On-device reactions

Oneira's expressive reactions run on the device, so its face can respond without relying on a cloud request.

See the local core ↗
Working prototype

Dual AI chip, OLED screen, and 1K camera

The working prototype includes a dual ai chip, oled screen, and 1k camera. The 1K camera is included in the $80 Founding Unit; its final production module may change before fulfillment.

Inspect the hardware ↗
Connected beta

Optional connected services

Optional compatible connected services can extend Oneira when you choose. Setup, permissions, internet access, provider terms, and usage costs remain separate. Connected experiences work only with compatible services the owner deliberately configures. Oneira does not connect automatically to every service.

See compatible connections ↗
Connected beta

Compatible connected tools

Compatible connected tools are optional and owner-configured. Nothing is preconfigured, and universal compatibility is not promised.

See connected features ↗
Web beta

Dream Big+ photo generation

The 1K camera supports deliberate, owner-triggered experiences. Optional Dream creation can transform a photo when the owner chooses it.

Open Oneira Create ↗
Developer preview

Focused helpers

Optional helper experiences have narrow, visible limits. Broader research, coding, and workspace actions remain in development.

Understand the agent path ↗

A simple flow

How it works in three steps.

  1. 01
    Request

    The owner begins a narrow note, checklist, or configured tool action

    →
  2. 02
    Route

    The app selects a supported owner-configured path

    →
  3. 03
    Review

    The owner checks output before any consequential extension

Controller job

Current preview tasks stay intentionally narrow.

Bounded preview tasks should do only what the owner explicitly requests and reject unavailable work instead of pretending it succeeded.

Browsing, coding, and workspace changes remain in development. The public product story does not expose private execution architecture.

Permission boundary

Do not turn one button into unlimited authority.

Connected agents and connected tools can reach files, messages, browsers, code, or business systems depending on configuration. Each added connection should have the narrowest credentials and scope possible.

Actions that spend money, publish, delete, send, reveal private data, or change external systems should require a clear confirmation and a reviewable destination.

Realistic support

The controller is only as reliable as the connected path.

Network outages, expired credentials, provider changes, service errors, tool downtime, and ambiguous requests can interrupt work. The interface should show failure rather than invent success.

Oneira is not a universal control surface for every external service. Optional connections depend on compatibility, individual permissions, owner configuration, provider terms, and availability that can change.

Agent control / intent before execution

A physical AI agent controller should control the permission boundary as carefully as it controls the task.

A physical AI agent controller is useful when it makes a narrow request feel deliberate and makes state visible without requiring a full workstation for every interaction. Oneira supplies a current on-device motion-reactive character layer and a Connected beta path through the app to optional compatible connected tools and services. A charming physical gesture should never imply unbounded authority over files, accounts, messages, money, or external systems.

The controller's real quality is visible control: the owner should know which service will receive the request, which permission applies, whether the task is waiting or blocked, and what requires confirmation. Connected support depends on compatible app setup, service behavior, credentials, terms, and availability that can change. Sight is a separate consent-led Connected beta visual path: an owner chooses one frame. Everyone present must agree before an owner-triggered Sight capture. Translation and creator capabilities remain in development rather than being assumed connected actions.

Decision framework

Four rules for an agent controller you can trust

  1. 01

    Keep the requested action narrow

    A physical control should map to a specific, understandable request instead of a vague instruction to "handle it." Define the task, expected destination, data scope, and acceptable result before connection. This prevents a compact interface from hiding the difference between a harmless lookup and a workflow that can affect people, systems, or money.

Questions, answered

The details that decide a purchase.

Does Oneira support bounded preview tasks?+

Yes. Current preview capabilities are intentionally narrow and should reject unavailable work rather than imply success.

Can it browse, code, or change a workspace today?+

No. Browsing, coding, and workspace changes remain in development and are not current included capabilities.

Can Oneira confirm dangerous actions automatically?+

Consequential actions should require explicit owner confirmation. A physical character should not be treated as blanket authorization.

What should be logged?+

The app should make the chosen service, requested action, permission, result, error state, and revocation path reviewable, subject to privacy and retention settings.

Keep exploring

Understand the product before you choose it.

Architecture guide / device to agentWhat is a physical interface for AI agents?→Connected beta / deliberate workflowsA deliberate start for a connected workflow.→Connected beta / optional toolsOptional tools, connected on purpose.→

Oneira / current availability

Preorder a Founding Unit · $80

500 Drop 01 units remain across 5 currently available colors. Availability is confirmed when secure checkout begins.

Preorder a Founding Unit · $80 →
oneira

Dual AI chip. OLED screen. 1K camera.

ComparePrivacyPreorder termsSupport
The honest boundary

Oneira's focused helpers have narrow, visible limits. Broader research, coding, and approved-workspace actions remain in development.

  • 02

    Require confirmation where consequences begin

    Sending, publishing, spending, deleting, changing records, sharing private data, or operating an external system should have a distinct confirmation step. A device animation, friendly personality, or prior success is not authorization. The app should make the consequential destination and proposed action reviewable before the agent or tool crosses that boundary.

  • 03

    Make state, logs, and errors reviewable

    An agent controller needs a truthful way to show waiting, routing, working, partial completion, failure, and cancellation. The small device can provide concise feedback, but the app should expose the chosen service, permission, relevant result, error, activity history, and revocation path. Reliable control includes knowing when the system did not succeed.

  • 04

    Design for changed services and revoked access

    Connected beta integrations can change as services update tools, terms, accounts, authentication, availability, or compatibility. Plan for expired credentials and stopped services before they occur. Use the least authority necessary, test read-only tasks first, keep recovery paths clear, and remove connections that no longer have a justified purpose.

    • Good fit

    Good fit if…

    • You want a tangible way to begin a small, configured agent task while retaining app-visible review and permission control.
    • You value on-device character expression but understand that optional connected tools require separately configured services.
    • You are willing to use least-privilege setup, confirmations, logs, and revocation rather than treating automation as hands-free trust.

    Choose differently

    Choose something else if…

    • Choose a desktop management surface if the workflow needs deep context, complex branching, or continuous operational oversight.
    • Choose a managed business automation product if a supported audit trail and vendor accountability are requirements.
    • Do not delegate consequential actions if you cannot define the permission, observe the result, and stop or revoke the workflow quickly.