# webAI App

Mac workspace for teams of Personas with their own roles, documents and tools. Local models can process tasks directly on the Mac.

Canonical profile: https://ki-agentenkompass.de/en/agents/webai/
Editor: [Michael Meierhoff](https://ki-agentenkompass.de/en/about/)
Sources reviewed: 2026-11-09
Basis: provider documentation. Examples are editorial suggestions, not hands-on product tests.

## Overview

This profile covers the public Mac app with Personas. The SDK and network are separate parts of the webAI platform.

- Provider: webAI
- Product family: webAI App
- Profile type: Builder for local agent teams
- Environment: Mac app with local models and optional connected services
- Connections: Includes Slack, DocuSign, Hugging Face and Apple apps
- Setup: Download the app, configure a Persona, assign documents and tools, and make the chosen model available.
- Pricing model: Local models without per-token billing. Check optional service prices separately.

## Key features

### Models on the Mac

The Mac app runs small reasoning models locally.

Source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Specialized Personas

Roles, knowledge and tools are selected for each Persona.

Source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Divide work between agents

Several Personas can contribute to the same task.

Source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Use documents with citations

Personas draw on assigned material and can cite it.

Source: https://www.webai.com/blog/expert-local-intelligence-in-action

### Connect tools and services

The app can connect services such as Slack and DocuSign.

Source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

## Example workflows

### Draft a proposal from project documents

- Your input: Project requirements, a rate card and three specialist Personas.
- Possible result: Draft with references to the documents used.
- What to check: Check prices, references and contractual commitments.

### Review outstanding signatures

- Your input: A DocuSign connection and a bounded date range.
- Possible result: A list of outstanding contracts and draft follow-ups.
- What to check: Confirm statuses in DocuSign and review drafts before sending.

## Requirements and limits

### Compatible Mac

Requirement: The public app is offered for Mac. Local models consume the device’s compute and memory.

Source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Connector access

Limitation: Connected services work with their assigned data and permissions. Local inference does not cover those external actions.

Source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Distinguish app and SDK

Limitation: SDK documentation describes additional interfaces. It does not establish which options are fully available in the desktop UI.

Source: https://www.webai.com/platform/sdks

## A good fit if …

- you want several roles to work from a common document collection
- you want local models for work on Mac files
- you want to review drafts from connected services before sharing them

## Less suitable if …

- you require a native Windows or Linux app
- you need documented recovery after machine failure for long automated runs

Editorial assessment based on provider documentation. Not a product test.

## Usable AI models

Fixed choice across providers

The Mac app provides local reasoning models. The platform documents model selection for Personas. Arbitrary external endpoints are unverified for this desktop app.

- Model families named: None named
- Model choice: You choose
- Your own API key: Not stated
- Router or custom endpoint: Not stated
- Local models: Yes

Partly documented · Sources accessed on: 2026-11-09

- [webAI App · Product introduction](https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges)

- [webAI App · Platform SDK](https://www.webai.com/platform/sdks)

## Europe & regional availability

Product access, country availability and data processing are assessed separately.

Reviewed scope: webAI App · Documented product access
Reviewed on: 2026-11-09

### Access and regional conditions

With requirements

The provider makes the Mac app publicly available. On-device processing and connected-service data paths should be assessed separately. A complete list of eligible countries is unpublished.

Region covered: EU/EEA, United Kingdom and Switzerland

Documented reason: The conditions of the specific access route and connected services apply.

Official source: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

## Accounts & permissions

Based on provider documentation. Permissions depend on the account, integration and organization policies. Review the permissions actually requested before use.

Configuration covered: webAI App · Account and permissions
Sources reviewed: 2026-11-09

### Account used

Documented with conditions: The platform identifies users, devices and Personas. Sign-in requirements for all Mac installations remain unclear.

Evidence: https://www.webai.com/platform/sdks

### Admin approval

Documented: Access and permitted actions can be configured per Persona. The provider shows drafts for approval.

Evidence: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Data access & team sharing

Documented: Connected apps and assigned documents determine a Persona’s reach.

Evidence: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

## Prices and included usage

Pricing checked: 2026-11-09

Local models without per-token billing. Check optional service prices separately.

The linked source does not fully specify plans or usage charges.

Pricing details: https://www.webai.com/platform/webai-app

## Operation & oversight

Based on public provider documentation. Assessments are editorial inferences from the stated properties. Actual time requirements and reliability have not been measured.

Deployment covered: webAI App · Execution and resumption
Sources reviewed: 2026-11-09

### Files and working state

Documented: Files reside on the Mac in local operation, and project context can be shared in Spaces.

Evidence: https://www.webai.com/platform/webai-app

### Sign-ins

Documented with conditions: Access to external services depends on configured connectors.

Evidence: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Starting further runs

Documented with conditions: Multiple Personas collaborate on one task. General scheduling for the public Mac app remains unclear.

Evidence: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Human intervention

Documented: Permissions are set per Persona and proposed messages can be reviewed.

Evidence: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Detecting errors

Open in reviewed sources: Reliable failure notifications are not specified in the product description.

Evidence: https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

### Resuming after failures

Open in reviewed sources: Resumption of tool actions after interruption is not documented.

Evidence: https://www.webai.com/platform/webai-app

### Editorial assessment

Editorial assessment for the documented access route. Each criterion lists requirements and unresolved operating details.

## Memory and conversation continuity

Scope: Persona knowledge, shared work context and platform chat memory
Sources accessed: 2026-11-09

Knowledge is assigned to Personas and workspaces. Detailed chat-memory rules are documented mainly in the platform API.

### Storage and scope

Partly documented: The app uses local files and reusable project context. The runtime API describes locally persisted chat memory with summaries, preferences and tool outcomes.

- [webAI App · Product overview](https://www.webai.com/platform/webai-app)
- [webAI App · Runtime API](https://developers.webai.com/api-reference/oasis-ai)

### Selection and timing

Partly documented: Users assign documents, skills and instructions to Personas. The runtime can save chats automatically when an app ID is present. How the public app extracts durable facts remains unclear.

- [webAI App · Personas](https://www.webai.com/blog/expert-local-intelligence-in-action)
- [webAI App · Runtime API](https://developers.webai.com/api-reference/oasis-ai)

### Inspect, edit and delete

Partly documented: The API offers an injection toggle and a method for clearing app chat memory. Equivalent Mac UI controls and editing individual memories remain unclear.

- [webAI App · Runtime API](https://developers.webai.com/api-reference/oasis-ai)

### Agent and channel boundaries

Partly documented: The app distinguishes Personas and shared Spaces. The runtime API filters chat memory by Persona or session ID. Full app access isolation is unverified.

- [webAI App · Product overview](https://www.webai.com/platform/webai-app)
- [webAI App · Runtime API](https://developers.webai.com/api-reference/oasis-ai)

### Retrieval

Partly documented: Assigned documents ground Persona answers, including citations. Ranking and selection of personal memories in the app are not described.

- [webAI App · Personas](https://www.webai.com/blog/expert-local-intelligence-in-action)
- [webAI App · Platform SDK](https://www.webai.com/platform/sdks)

### Session reset and compaction

Partly documented: Project context can be reused. The runtime describes chat context across sessions, while resumption of interrupted app tool actions remains unclear.

- [webAI App · Product overview](https://www.webai.com/platform/webai-app)
- [webAI App · Runtime API](https://developers.webai.com/api-reference/oasis-ai)

## Context

The public app is offered for Mac. Local models consume the device’s compute and memory.

Official product documentation reviewed. Workflows and assessments are editorial; hands-on product tests are outside the scope of this directory.

## Sources

- https://www.webai.com/platform/webai-app

- https://www.webai.com/blog/webai-launches-webai-app-for-on-device-ai-with-no-per-token-charges

- https://www.webai.com/blog/expert-local-intelligence-in-action

- https://www.webai.com/blog/how-to-bring-your-people-and-ai-together-with-spaces

- https://www.webai.com/platform/sdks

- https://developers.webai.com/api-reference/oasis-ai

## Provider

webAI develops a platform for models and agents on hardware you control.

- Operator: webAI
- Organization: Software provider
- Website: https://www.webai.com/platform/webai-app
- Reviewed: 2026-11-09

- Provider source: https://www.webai.com/platform/webai-app

[Sources and method](https://ki-agentenkompass.de/en/sources/)
