Claude Sonnet 5 is now available in Babbily.

Sonnet 5 brings Anthropic’s newest frontier model into the Babbily model picker, giving teams a stronger default for deep reasoning, writing, coding, analysis, and complex multi-step work. Sonnet 4.6 is still available, but it has been moved into sunset status as Sonnet 5 becomes the recommended Sonnet model.

Also added:

• GLM 5.2 Fast as a fast-mode option for GLM 5.2

• Nano Banana 2 Lite for lighter-weight image generation

• Updated model ordering and catalog metadata so the selector reflects the newest recommended models first

Babbily now supports passkeys for faster, passwordless sign-in on supported browsers and devices.

You can add a passkey using Face ID, Touch ID, Windows Hello, a password manager, or a hardware security key. Once added, you’ll see Sign in with passkey on the web login screen.

New users may be prompted to add a passkey during setup, and existing users can add, rename, or remove passkeys anytime from Settings > Account.

Passkeys are currently available on web browsers. Native app passkey support will come later.

Babbily now includes a full Finance area for market research, company pages, earnings, prediction markets, and watchlists. News also has richer discovery, source-backed article generation, local context, and market highlights in the side rail.

What’s new

Finance in Babbily.

Open Finance to review US markets, crypto, earnings, prediction markets, market summaries, company pages, price charts, key issues, analyst context, peers, and historical data where available.

Watchlists and company research.

Add companies to a watchlist, open company pages from market rows, and ask follow-up questions from a Finance page with the page context already attached.

Richer News discovery.

News now supports topic tabs, location-aware discovery, source citations, generated article views, article follow-up chat, weather, market outlook, and trending company links.

Better performance and fallback behavior.

News and Finance data now refresh more smoothly, keep recent results available during temporary provider issues, and load common market and news panels faster.

We’ve added three powerful new models to Babbily: Sakana Fugu Ultra, Kimi K2.7, and GLM 5.2.

This update gives Babbily users even more choice across reasoning, coding, long-context work, and agentic task execution.

What’s new

Sakana Fugu Ultra

Sakana Fugu Ultra is now available in Babbily.

Fugu Ultra is different from a traditional single-model experience. Instead of relying on one model to answer every prompt, it coordinates multiple frontier models, routing work to different agents depending on the task and combining the results into one final response.

This makes it a strong option for complex prompts that benefit from multiple perspectives, stronger reasoning, and deeper task review.

Best for:

• Advanced reasoning

• Scientific and technical analysis

• Code review

• Complex multi-step prompts

• Comparing and synthesizing multiple approaches

Kimi K2.7

We’ve also added Kimi K2.7, a coding-focused model built for long-context software engineering tasks.

Kimi K2.7 is designed to follow instructions more reliably in long contexts, perform better on extended coding tasks, and support agentic workflows where the model needs to reason through multiple steps.

Best for:

• Coding

• Debugging

• Multi-file reasoning

• Long-context development work

• Agentic software tasks

GLM 5.2

GLM 5.2 is now available in Babbily as well.

GLM 5.2 is built for long-horizon software engineering and agentic coding workflows. With a large context window, it is especially useful when users need to work across larger projects, longer files, deeper instructions, or extended planning sessions.

Best for:

• Long-context coding

• Repo-scale reasoning

• Multi-step engineering workflows

• Refactoring support

• Technical planning

Why this matters

Babbily is built to give users access to the best AI models in one place. With Sakana Fugu Ultra, Kimi K2.7, and GLM 5.2, users now have even more flexibility to choose the right model for the task.

Whether you’re writing, coding, researching, analyzing, planning, or building, these new models expand what you can do inside Babbily.

They are available now.

We’re rolling out a new Canvas experience in Babbily, giving you a dedicated workspace to create, edit, and refine AI-generated work without leaving the chat.

Instead of copying content into another doc, spreadsheet, or editor, you can now work directly inside Babbily’s Canvas. Create drafts, edit text, adjust tables, work with structured outputs, and continue prompting the model with your latest edits already included as context.

What’s new

Edit directly in Canvas
AI-generated work can now open in a durable side panel where you can make changes directly. Whether you’re working on a document, table, code artifact, or structured output, the Canvas becomes the live workspace for that response.

Your edits stay in context
When you edit something in the Canvas and continue prompting, Babbily shares the current Canvas state with the model. That means follow-up prompts like “make this shorter,” “turn this into a table,” or “update the second section” are based on the latest version you edited, not just the original chat message.

JSON Rendering
Babbily can now render structured JSON outputs into cleaner, more usable visual components. This makes generated dashboards, tables, callouts, metrics, and other structured outputs easier to view and work with.

Download your work
Canvas artifacts can be exported in useful formats depending on the artifact type, including Word, PDF, CSV, Excel-ready spreadsheet files, and JSON.

Push directly to connectors
You can send Canvas work directly into connected tools through Babbily connectors. This makes it easier to move from AI-generated output to the places where your team actually works, without extra copy/paste steps.

Canvas mode is a big step toward making Babbily feel less like a chat box and more like a real AI workspace.

Claude Fable 5 is now available on Babbily — with a 1 million token context window.

Fable is a Mythos-class model built for long-running, complex, and asynchronous work, with robust safeguards designed for serious tasks.

That means fewer interruptions, less context loss, and more room to think through big projects from start to finish.

Available now on Babbily.

We’ve launched the new Babbily documentation site at https://docs.babbily.com.

This gives users, teams, and admins one central place to learn how to use Babbily, roll it out across a workspace, manage billing and privacy, troubleshoot issues, and get more from AI.

What’s new

A central home for Babbily help

The new docs site brings together product guidance, setup instructions, troubleshooting, privacy details, team administration, and feature education in one searchable hub.

Faster getting started

New users can start with Quickstart, Account Access, Glossary, Role-Based Getting Started, Task Recipes, and Prompting Patterns to learn the basics faster.

Better guidance for teams and admins

Team owners and admins now have dedicated docs for team setup, rollout checklists, admin workflows, offboarding, access reviews, procurement, security, privacy, and data controls.

Clearer AI feature education

The docs explain how to use Babbily’s core AI capabilities, including chat, model selection, file attachments, deep research, search, visual replies, translation, voice dictation, image and video generation, memory, saved prompts, tools, skills, and connectors.

Billing, usage, and limits in plain English

We added clearer documentation for plans, monthly API usage budgets, billing changes, plan FAQs, known limits, context windows, supported environments, browser support, file support, and service status expectations.

Built-in troubleshooting paths

Users now have dedicated troubleshooting guides for login issues, invites, connectors, common errors, degraded service behavior, device support, and contacting support.

Changelog now lives in docs

We also added a changelog section so users can track what’s new, what changed, and what was fixed in Babbily.

Why it matters

Babbily is becoming more than a chat app. It is an AI workspace where users can work across models, files, tools, memory, apps, media, and teams.

The new docs site makes that workspace easier to understand, easier to adopt, and easier to support.

We’ve updated how Babbily tracks AI usage across chat, research, media generation, memory, and connected tools.

This release gives users a more consistent usage experience across the platform, with plan limits now managed through Babbily’s AI Gateway usage system instead of the older token balance flow. The goal is to make usage tracking more accurate, easier to maintain, and better aligned with the real cost of running advanced AI models and media tools.

Team plans also received improvements so usage is handled more cleanly across accepted team members. This helps keep shared team access more reliable as members join, leave, or are added to a workspace.

We also made backend improvements to usage refresh timing, plan checks, and media generation support. These changes help Babbily handle high-cost AI features more responsibly while keeping the user experience smooth across chat, image generation, video generation, research, translation, and other AI-powered workflows.

This update also simplifies parts of the platform by removing older usage-tracking paths and standardizing more AI activity through the same gateway-backed system.

We’ve released a major Babbily platform update focused on making chat more powerful, more connected, and easier to control.

Babbily now supports a new tool system that powers web search, deep research, site mapping, finance research, weather, stats, QR codes, image generation, video generation, memory, skills, and connected tools through Smithery MCP connectors. Users can also choose how much Babbily should use tools with new Auto, Manual, and Off modes.

This release also improves the live chat experience. Tool progress, research status, memory citations, connector diagnostics, and media generation states now stream more clearly so users can better understand what Babbily is doing while it works.

Memory and profile personalization have been upgraded as well. When memory is enabled, saved profile details can now sync into Babbily’s memory system, helping the assistant better understand user preferences and work style over time.

We also added saved-thread context compaction, which helps long conversations stay useful by preserving important context while reducing unnecessary thread bloat.

Additional improvements include a cleaner floating thread header, better attachment and connector persistence in conversations, improved reliability for long-running research and tool jobs, and backend upgrades to support future connected workflows.