Wren AI
Open-source text-to-SQL agent that turns questions into dashboards
Overview
Wren AI is an open-source GenBI agent answering natural-language questions across 20-plus data sources, including BigQuery, Snowflake and Postgres, by grounding SQL generation in a governed semantic layer. Usable via UI, API or MCP; self-funded, with cloud and self-hosted tiers atop the free core.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- 2 projects, 2 members
- 10 tables per project
- 20 monthly credits
- 13,200 annual credits
- Unlimited projects, dashboards, spreadsheets
- Embedded AI API and dbt integration
- 24,000 annual credits
- Row- and column-level data control
- MCP integration, advanced RBAC, custom SSO
- Business tier from 5 concurrent sessions
- Enterprise Plus from 10 concurrent sessions
- On-premise and air-gapped support
What you can produce with Wren AI
- Natural-language querying across 20+ data sources
- Governed semantic/context layer for grounded SQL
- Dashboards, spreadsheets and traceable workflows
- MCP client support for AI agents
- Self-hostable via Docker or managed cloud
- dbt integration and embedded AI API
ASEAN Perspective
Wren AI in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Wren's angle on text-to-SQL is the "context layer" — rather than firing a raw question at an LLM and hoping the schema is self-explanatory, it grounds queries in defined business semantics, which matters once you're past a demo database and into a real warehouse with ambiguous column names and multiple tables that could plausibly answer the same question.
It's bootstrapped: no venture funding, a reported $1.5 million ARR, built by a roughly 14-person team, and it has genuinely grown its open-source following — GitHub stars roughly quadrupled through 2025 to 13,000-plus. Cloud pricing jumps hard between tiers, from free to $179/month Essential to $559/month Enterprise, and self-hosted plans are metered by concurrent sessions rather than seats, which takes some getting used to. For teams that want agent-queryable BI without handing a closed SaaS vendor their warehouse credentials, it's one of the more credible open options next to closed tools like Hex or Julius.
What people say
Wren AI's core claim is "GenBI," generative business intelligence where both humans and AI agents can query a warehouse in plain English and get back grounded SQL, not just a plausible-looking guess. The mechanism is what the company calls an open context layer: a semantic model that maps business terms to actual schema before generation happens, which is the difference between a demo that works on a toy database and something that survives contact with a real Snowflake instance full of ambiguously named columns.
The open-source project (Canner/WrenAI on GitHub) has real momentum: it started 2025 with roughly 3,000 stars and closed the year past 13,000, a genuine acceleration rather than a one-time launch spike. The company's own year-in-review states it moved from "open source to agentic BI in production" over that period, and claims 10,000-plus cloud users on its commercial plans. Unlike most of this category, Wren AI has taken no outside venture funding — it's self-funded and reports around $1.5 million in annual revenue with a team of about 14, up from 8 in 2024, worth noting as a durability signal in a space where several well-funded competitors have folded or pivoted.
Pricing has a wide gap between tiers: a genuinely usable Free plan (2 projects, 10 tables, 20 monthly credits) sits well below Essential at $179/month (unlimited projects and dashboards, embedded API, dbt integration) and Enterprise at $559/month (row and column-level access control, MCP integration, custom SSO) — there's no mid-tier softening that jump for a team that's outgrown Free but doesn't need Enterprise governance. Self-hosted deployment is licensed differently again, by concurrent session count rather than seats or spend, starting at 5 concurrent sessions for the Business tier and 10 for Enterprise Plus (required for OEM/resell or regulated deployments). Data-source coverage is broad — 20-plus sources including BigQuery, Snowflake, Postgres, ClickHouse, Redshift and Databricks — wider than most open-source text-to-SQL alternatives.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Wren AI's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Wren AI unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Wren AI directly →
Spotted something out of date? Suggest an update →
Wren AI in the news
More in Research & Data
