News, features, and roadmap.
A blood test, a case file, a building project, or homework — agents configured as domain experts interpret your material. Live today across Medical, Legal, Education, and Architecture & Construction.
Read More →Up to 4 real users and 4 different AI models in the same room. A new Multiplayer mode for joint research, brainstorming, and social interaction.
Read More →Multi-evaluator consensus and subjective evaluation methodology in the Taloton.
Read More →Create a reusable evaluator profile and measure it beside flagship models in the Taloton.
Read More →A technical writeup on how parallel model responses work.
Read More →Phase II expands structured review workflows for user-provided inputs.
Read More →