Artificial Analysis: Keenable Search 1.0 takes 1st and 2nd on its Search Index

Artificial Analysis says Keenable Search 1.0 ranks first (pro mode, 83) and second (realtime mode, 82) on its Search Index, ahead of Exa (auto, dynamic highlights) and Perplexity (medium), both at 80. It ran both modes in its Stirrup agent harness on 1,700 search-focused tasks, with GPT-5.6 Luna (medium) as the model.

• Cost: Keenable charges $4 per 1,000 searches in both modes, which Artificial Analysis puts at about $0.040 per task, over a third below the board median of $0.063.
• Speed: 15.0 seconds per task in realtime mode and 15.5 in pro mode, 2nd and 3rd on the board.
• Other benchmarks: both modes score 88 F1 on DeepSearchQA, 1st and 2nd of 26. Pro mode also ranks 1st on BrowseComp (87%) and AA-Omniscience (73%).

These are Artificial Analysis's own measurements.