Perplexity's on-device parser scores 65% on ParseBench, nearly double Hermes
llama_index · x · 2026-08-27
Perplexity shared benchmark results for its Computer product: parsing runs entirely on-device, so sensitive documents never leave the machine. On ParseBench-100, Computer scores 65.1% versus 34.6% for Hermes and 13.9% for Pi — while also finishing in the least time with the fewest tokens.
More from Apps
- Radar Searches 130k+ Transcribed Podcasts, Tracks Trends with Entity Extraction — Scobleizer · 2026-08-27
- Finite 3D tool showcases complex hand-made part construction — jakedahn · 2026-08-27
- Analyst Agent launches: AI for raw analytical chemistry instrument files — ycombinator · 2026-08-27
- Viral AI assistant Instinct raises $350M total at a $2.5B valuation — KateClarkTweets · 2026-08-27
- Sarvam AI's speech-to-text and foundation models show maturity — abhish18 · 2026-08-27
- Grok Bot now available to SuperGrok and Cursor Pro users — XFreeze · 2026-08-27