ChatLLM Teams by Abacus AI lets you schedule recurring tasks such as research summaries, reports, and reminders. This guide retains its February 2025 video and screenshots; they show the interface at that time. Public documentation checked: October 4, 2026. The update below reflects current published instructions. No authenticated account, task run, email delivery, or connector was tested for this update.
Introduction to ChatLLM Teams
ChatLLM Teams combines Large Language Models (LLMs) with tools for web search, image generation, code, and document analysis. The original guide named GPT-4o, Sonnet-3.5, and Gemini 2.0; those links are historical model references, not a current availability list. Check the model picker and Abacus’s current FAQ for the models available to your account.
As checked on October 4, the Billing FAQ lists Basic at $10 per user per month with 20,000 credits, and Pro at $20 total per month with 30,000 credits. Basic includes limited Abacus AI Agent access; Pro removes those access restrictions while credits remain available. First-month promotions can reduce the included credits. These are subscription terms, not a guarantee of unlimited task runs: check your account’s Billing and Credit Usage pages before enabling recurring work.
Unveiling the “Tasks” Functionality
Tasks run on a schedule you specify. Abacus documents recurring research, reports, and reminders, plus Agent Tasks that can use connected services, databases, or browser workflows. External actions need the relevant access and permissions; a prompt alone does not connect an account.
Setting Up Tasks
The ChatLLM Tasks FAQ describes selecting Task under More near the search bar. The Agent Tasks guide instead points to Tasks in the left navigation. Use the path shown in your interface and describe the work in natural language. For example: “Every weekday at 9 AM in America/Vancouver, find one newly published AI paper, link the original paper, and summarize its findings and limitations.” Check the configured time and recurrence. Run the task manually once and inspect its sources and output before relying on the schedule.
Customizing Task Schedules
You can request daily, weekly, or monthly work and enable email alerts for task runs. The Tasks FAQ says each task has a conversation containing its runs; the Tasks list lets you edit, pause, resume, or delete it. Check that the first scheduled run appears there and that an enabled alert reaches you. Tasks consume credits and have rate limits; review the Agent Tasks guide’s current Usage Limits and your billing usage. Pause unnecessary tasks rather than leaving every schedule active.
Practical Applications
These are examples of work you can request, not results tested for this update:
- Academic Engagement: Request a daily or weekly paper digest with original publication links, dates, and a short explanation of why each paper is relevant.
- Travel Planning: Request an itinerary for specified dates and interests. Check opening hours, availability, and prices at the original sources before booking.
- Professional Development: Request an industry-news or learning-resource digest with source links and dates. Define which topics and sources matter to you.
- Personal Management: Request recurring reminders or a weekly activity plan. Specify the time zone, frequency, and end date, then check the task’s schedule.
Integration with Other Features
For a workflow that reads email, updates a database, or posts to another service, use the documented Agent Tasks workflow and connect only the services it needs. Abacus lists user connectors, MCP servers, OAuth, APIs, and secrets management. Verify the connector can perform the intended operation with controlled data; a “connected” label alone is insufficient. For Self-Improving Tasks, define an objective, a measurable baseline, permitted changes, and a repeatable evaluation. Regular recurring work does not require that extra improvement loop.
Security and Data Privacy
The Data Security and Privacy FAQ says Abacus does not train on customer data, has provider agreements protecting it from model training, and encrypts data. It also states enterprise compliance including SOC-2 Type-2 and HIPAA. These are the vendor’s statements, not an independent security audit by Kingy.ai. Review your organization’s requirements before connecting sensitive data, and check whom you have shared conversation URLs with.
Before Leaving a Task Running
Start with one task and a modest frequency. Confirm the prompt, schedule, sources, intended recipients, and permissions. Inspect one manual run, then check the first scheduled run and its credit use. If the result or delivery is wrong, pause the task and correct its instructions or access before resuming.
For a research digest, require original-source links and dates rather than an unsupported summary. For an action that changes records or sends a message, test on controlled data and verify the actual result. The task history is where to check what ran.
Platform Guides
Presentation Workflows
Productivity Workflows
The Kingy Brief
Get The Kingy Brief.
AI changes, original tests and one practical thing to try. Fridays at 09:00 Vancouver time.
Free · Double opt-in · Unsubscribe anytime
Signup help and newsletter schedule
Signup form provided by Beehiiv. After submitting, check your inbox for "Confirm your subscription to The Kingy Brief" and open its confirmation link. Check Spam or Promotions if you cannot find it.
Fridays at 09:00 Vancouver time: source-checked AI changes, original tests and one practical thing to try. The weekly restart begins October 9, 2026. We skip a week when there is not enough verified material. Free. Unsubscribe anytime.