Sonar Reasoning Pro (Thinking): benchmark results

Provider: Perplexity. Access: API.

Unified ELO 1570 ± 1, rank #840 of 3078 rated models, from 10 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Tinybird AI SQL Benchmark - First-Attempt Success Rate98Questions answered with a valid query on the first attempt (78.1
Tinybird AI SQL Benchmark - Success Rate100Questions answered with a valid query within 3 retries (%)72.5
Tinybird AI SQL Benchmark - Exactness49.13Result exactness vs human reference queries (0-100)56.4
DeepResearch Bench - Citation Accuracy45.19Score (%)16.7
DeepResearch Bench - Effective Citations9.39Effective Citations16.7
DeepResearch Bench - Readability42.42Score (%)15.9
DeepResearch Bench - Instruction Following44.93Score (%)13.6
DeepResearch Bench37.76Score (%)11.4
DeepResearch Bench - Insight31.65Score (%)11.4
DeepResearch Bench - Comprehensiveness34.96Score (%)9.1

Interactive version: theaggregate.ai/model?slug=sonar-reasoning-pro-thinking · How It Works · Data refreshed daily, snapshot 2026-09-19.