VivaDictaVivaDicta
← All releases

2.2.1

September 28, 2026

Four new AI models for enhancement - GPT-6 Sol, GPT-6 Luna, Claude Opus 5.5 and Grok 4.7 - each measured on the VivaDicta benchmark before it joined the picker.

New AI models

  • Grok 4.7 - quality 9.7 at 2.5 s, well over twice as fast as Grok 4.6 at the same quality.
  • Claude Opus 5.5 - quality 9.6 at 2.4 s.
  • GPT-6 Sol - quality 9.5 at 1.05 s, and GPT-6 Luna - quality 9.4 at 1.5 s.
  • The model ranking on the AI Providers screen includes all four.