

MinusX
β Text source-checked Nov 1, 2026Open source agentic business intelligence platform designed for founders.
Source-checked against the vendor's site Β· rebuilt from the vendor's own pagesProfile updated Link checked Pricing captured How we rate
In short
MinusX is an open-source agentic BI platform that automates SQL analysis, dashboards, and boardroom-ready reports. It supports self-hosting or managed cloud with BYO LLM keys.
The vendor says that MinusX is an open source agentic business intelligence platform that ranks first on DataAgentBench, a benchmark for realistic data tasks from UC Berkeley's EPIC lab. The platform staffs the data stack with agents that handle SQL analysis, dashboards, metrics, alerts, and reports.
MinusX converts analysis into boardroom-ready slide decks or scrollable data stories, including charts, annotations, and takeaways. Users can build rich visualizations such as maps and funnels, and tweak details in a full SQL editor or point-and-click query builder before adding charts to dashboards.
The agent auto-learns business context from interactions, and users can test its performance using built-in evaluations.
MinusX connects to data sources including Postgres, BigQuery, Athena, DuckDB, Google Sheets, and CSV uploads. The tool integrates with Slack for asking questions and exposes an MCP server for use by other AI agents.
Users can self-host the open source version in their own environment or use the managed MinusX Cloud service. The vendor says that the cloud platform supports bringing your own LLM keys, which reduces the cost of any plan by 50%.
According to MinusX: The platform staffs the data stack with agents that handle SQL analysis, Dashboards, metrics, alerts, and reports. Its pages list this strength: The core software is source-available and inspectable.
β The AI Tool Bible editorial team
Pros
- β The vendor says that ranks first on DataAgentBench, a realistic benchmark for data tasks from UC Berkeley's EPIC lab.
- β The core software is source-available and inspectable.
- β Supports connections to Postgres, BigQuery, Athena, DuckDB, Google Sheets, and CSV files.
- β Includes Slack integration and an MCP server for AI agents.
- β Features an auto-learning knowledge base and evaluation tools for testing agent accuracy.
Pricing (captured )
| Tier | Monthly | Annual | Key features |
|---|---|---|---|
| Open Source | Free | Free |
|
| Enterprise | Custom | Custom |
|
Prices change; we show the grid as captured on Oct 12, 2026. Check current pricing on the vendor's site β
Facts
- Platforms
- web, slack, api (not source-checked)
- API available
- Yes (not source-checked)
- Free trial
- Yes
- Open source
- Yes
- Model used
- Claude (BYO LLM key supported) (not source-checked)
- Hacker News
- 6 stories Β· top 469 points Β· last Aug 23, 2025 Β· checked Nov 16, 2026
- Earliest Wayback capture
- 2024-07-21 (earliest archived capture of the site, not a launch date)
Where this tool sits
Use cases
Frequently asked
- How much does MinusX cost?
- The Open Source plan is free forever. The Founders plan costs $40 per user per month with no platform fee. The Team plan costs $600 per month and includes 20 users. Using your own LLM keys reduces cloud plan costs by 50%.
- Which data sources does MinusX integrate with?
- MinusX connects to Postgres, BigQuery, Athena, DuckDB, Google Sheets, and CSV uploads. It also integrates with Slack for asking questions and exposes an MCP server for use by other AI agents.
- Can I self-host MinusX?
- Yes, users can self-host the open source version in their own environment. Alternatively, you can use the managed MinusX Cloud service, which supports bringing your own LLM keys to reduce costs.
- What types of outputs can MinusX generate?
- The platform converts analysis into boardroom-ready slide decks or scrollable data stories, including charts, annotations, and takeaways. It also supports building rich visualizations like maps and funnels, as well as scheduled reports and proactive alerts.
- How does the AI agent learn my business context?
- The agent auto-learns business context from user interactions. Users can test its performance using built-in evaluations. The platform ranks first on DataAgentBench, a benchmark for realistic data tasks from UC Berkeley's EPIC lab.