SQL

Natural language to SQL query generation evaluates text-to-query fidelity and schema reasoning. This task is particularly relevant for analytics chat assistants and simplified database interfaces where users need to query data using natural language. Models must understand both the intent behind the question and the structure of the underlying database schema.

Duration

20s

Input Tokens

1039

Output Tokens

1128

Cost

$0.00

Context

Input

Hi I'm Oscar that enrolled in 2018, I don't remember what courses I completed, can you help me?

Expected output

SELECT DISTINCT c.title FROM users u JOIN enrollments e ON u.user_id = e.user_id JOIN lessons l ON l.course_id = e.course_id JOIN completions cp ON cp.lesson_id = l.lesson_id AND cp.user_id = u.user_id JOIN courses c ON c.course_id = l.course_id WHERE u.name ILIKE '%oscar%' AND DATE_TRUNC('year', e.enrolled_at) = DATE '2018-01-01';

Model output

SELECT T2.title FROM enrollments AS T1 INNER JOIN courses AS T2 ON T1.course_id = T2.course_id INNER JOIN users AS T3 ON T1.user_id = T3.user_id WHERE T3.name = 'Oscar' AND T1.status = 'completed' AND EXTRACT(YEAR FROM T1.enrolled_at) = 2018;