Help Center

Abkus AI turns a plain-language goal into an executed project. You describe the outcome; the app assembles a team of specialized AI agents, splits the work into sprints and runs it end to end.

1. Sign in with your Abkus account (single sign-on across every Abkus app).
2. Create a project to hold related missions.
3. Describe your first mission and let the team run.
4. Review the delivered output and export it.

A project groups missions that belong together. A mission is one autonomous execution run: brief in, finished deliverable out — it can span multiple sprints and dozens of agent tasks.

• Missions run in the background; you can follow their progress live.
• Each mission keeps its own team, resolved once and reused across chained steps.
• Finished missions can be exported as Markdown, and your mission list as CSV.

Every mission is executed by a team of agents with distinct roles — an orchestrator plus specialists. Teams are chosen by domain (software, legal, medical, business, content, research, operations), so the right specialists work on the right problem.

• You can create your own teams and reuse them across missions.
• Agents run on models selected by the Abkus Hub — you never manage provider keys.
• Elite plans allow larger teams and deeper context per mission.

Templates give a mission a proven starting structure (for example a systematic review). Schedules let a mission repeat automatically on a cron expression, so recurring work runs without you.

The chat works in three modes: conversation (ask anything), planning (structure work before committing) and mission (turn the conversation into an executed mission). Just write naturally — the app detects the intent.

Organizations let you group projects and share agents across a team. Each organization has its own agents you can attach to missions.

Every plan includes a monthly mission allowance. Unused missions roll over within your plan limits. If you need more in a cycle, add-on mission packs are available without changing plan.

Billing is handled entirely by the Abkus Hub — one account for every Abkus app. Plans and invoices live there, and every subscription carries a 14-day money-back guarantee.

• Active or trialing: full access.
• Payment fails (days 1–14): full access continues with reminders to update payment.
• After 14 days: read-only — all your data stays visible and exportable, writes pause until you reactivate. Reactivation is instant after payment.
• Cancelled: read-only retention window, export always available.

Change language and appearance from the header. Account details, password and security are managed in the Abkus Hub account area. No AI provider key is needed — the Hub provides the models for every plan.

Your data is yours. Download a finished mission as Markdown from the mission view, or export your whole mission list as CSV. Export keeps working even in read-only mode.

• Do I need an AI API key? No. The Abkus Hub provides the models on every plan; there is nothing to configure.
• What counts as a mission? One autonomous run from brief to deliverable, including all its sprints and agent tasks.
• Can I cancel anytime? Yes — from the Hub billing area, with no penalty.
• What happens if a payment fails? You keep full access for 14 days with reminders; after that the account becomes read-only (never locked out) until reactivation.
• How many languages does the app support? The interface ships in 14 languages.

Elite plans can call the Abkus AI API directly using a Hub-issued access token. Missions, projects and chat are available as JSON endpoints.

• Authenticate with the Hub bearer token (no app-specific keys).
• Create a mission, poll its status, and fetch the delivered output.
• Outbound webhooks are not available yet — poll the mission status instead.