Jayant Arora

About

Software engineer and tech explorer. I test new tools and products of all kinds, and hunt the best ones for the community.

Badges

Tastemaker
Tastemaker
Tastemaker 10
Tastemaker 10
Tastemaker 5
Tastemaker 5
Gone streaking 10
Gone streaking 10
View all badges

Forums

•

6d ago

Fish Audio Live 3 - Voice agents with conversational timing built into the model

Fish Audio Live 3 is a full-duplex voice model that keeps listening while it speaks, so pauses, “mm-hmm” acknowledgments and changes of mind shape what it does next. Per Fish Audio, timing is learned from real human conversations inside the speech model, not a pipeline of separate parts, and it can decide to respond in as little as 186 ms. English only at launch. On the Fish Agent Platform and the API, with on-prem for enterprises.
•

7d ago

Claude Fable 5.2 and Mythos 5.2 - Anthropic’s top model, now up to 30% faster than 5.1

Claude Fable 5.2 and Mythos 5.2 are Anthropic’s upgraded frontier models for the hardest technical work. They are the same underlying model: Fable is open to all users, while Mythos, without some of Fable’s cyber and biology safeguards, goes to vetted organizations. Per Anthropic, outputs are up to 30% faster than 5.1 at a lower price ($8 input, $40 output per million tokens). Fable 5.2 scores 49.1% on FrontierCode 2.0, up from 32.4%. API ID: claude-fable-5-2.
•

8d ago

BioBasicsBench by OpenAI - A benchmark for how AI agents read raw lab data

BioBasicsBench is OpenAI’s benchmark, developed with enzyme-design company Ligo, for how AI agents handle everyday wet-lab science. It has 100 tasks that scientists built from real experiments, covering gels, colony plates, microscopy, sequencing reads and more, graded against expert rubrics. Per OpenAI, GPT-6 Astra scored 66.7% at its highest effort setting, and higher scores do not yet amount to reliable laboratory judgment.
View more