Agentsql vs Natural Language SQL

AgentSQL is a web service that turns plain-English questions about connected databases into read-only SQL queries, then returns a chart, table, and plain-English answer. Each response shows the exact SQL, and users can rephrase their question to have the service rewrite and rerun it. Supported connections are Postgres, MySQL, Snowflake, and BigQuery. Connections use least-privilege read-only credentials, can be limited to selected tables, and can be revoked. AgentSQL says it encrypts connections in transit, does not copy database tables or build a shadow warehouse, and does not use database contents, queries, or results to train AI models. It stores query history; Team and Scale also retain result rows. Plans are billed yearly: Starter costs $24 per month, Team $59 per month, Scale $179 per month, and Enterprise $299 per month. Monthly allowances are 600, 1,500, 5,000, and 8,000 questions respectively, with no more than one tenth usable in a day. The service targets founders, operations teams, analysts, and product managers.

Natural Language SQL turns plain-English questions about databases into SQL, validates and runs the query, then streams back results. It connects to PostgreSQL and MySQL, including multiple databases of the same type in parallel. Hybrid retrieval combines BM25 and sentence-transformer vectors to select schema context for a question. Users can review and edit generated SQL before execution, while the interface shows progress through query stages. Results can be sorted, paginated at 50 rows per page, downloaded as CSV or JSON, or copied as tab-separated text. SELECT queries receive a default limit, explicit result counts are capped at 1,000 rows, and execution times out after 30 seconds by default. INSERT, UPDATE and DELETE are allowed by default, with warning banners; the project recommends a read-only database user to block writes. Ollama runs locally by default, keeping questions, schema and generated SQL on the user’s machine. OpenAI, Gemini and Groq are optional cloud providers and receive the question, relevant schema portion and SQL context. The project is free under GPL-3.0. Its recommended Docker setup needs Docker 20.10 or later with Compose v2 and about 10 GB of free disk space.