News
News
What we’re shipping, from new models to new products, straight from the Babbily team.
Latest announcements
New models, features and improvements as they ship.
Improvement
Auto, re-tuned for the newest models
We tested how Auto handles each kind of task against the newest models, comparing answer quality, speed and cost. Auto now sends more of your work to the models that came out ahead.
Models
Claude Fable 5.2, two more new models, and Ultrafast
Claude Fable 5.2, Gemini 4 Carbon and Muse Ember join the model picker, DeepSeek V4.1 Flash 1017 gets an even quicker speed, and video generation moves to a new model that adds sound.
Release
Voice conversations, visuals in chat, and a better memory
You can now talk with Babbily out loud and hear it answer, and its replies can show charts, diagrams, maps and interactive explainers right where they help.
Release
Meet Auto, a calmer Babbily, Claude Sonnet 5.5, and GPT-6.1 Sol
Auto now picks a model and reasoning level for each message, the app has a quieter design, and Claude Sonnet 5.5 and GPT-6.1 Sol join the picker.
Models
Claude Opus 5.5, GPT-6 Sol and Luna, and Grok 4.7
Four new chat models arrive in the picker, with Fast mode on three of them. Several older models move to Sunsetting.
Improvement
Attach any supported file to any chat model
Every chat model now accepts documents, code, images, audio, video, and ZIP archives. Babbily reads files the selected model cannot, and reads each file only once.
Improvement
Replies stream in as Babbily writes them
Chat replies now stream into the thread as they are written, and started about 47% sooner in our testing.
Improvement
More relevant pictures from web search
Web search adds pictures only when they fit the question, and leaves out logos, ads, and tiny thumbnails.
Models
GPT-6 Astra, updated image generation, and more control over ongoing work
GPT-6 Astra and GPT Image 2.5 Flare arrive, threads can move between projects, and memory gets a review queue.