Anthropic is expanding Claude from a chatbot into a work platform with spreadsheets, whiteboards, meeting functions, and embedded partner apps. The additions are aimed mainly at teams that spend time planning, analyzing data, or moving results between multiple programs. At the same time, the new Claude Fable 5.2 and Mythos 5.2 models are intended to handle more demanding work, although their performance, price, and access differ considerably.
From chat window to work platform
The expansion centers on Claude Sheets, Claude Whiteboards, Claude Data, and Claude Meetings. According to the overview of Claude’s new workplace features, Sheets lets you create and edit spreadsheets with Claude, including formulas and multiple tabs. Whiteboards provide a shared visual space where a team can collect ideas and ask Claude to organize or summarize them.
Claude Data, previously called Dashboards, connects to services such as Snowflake and Salesforce. Claude can use them to produce dashboards and data visualizations. The feature is generally available on paid plans, according to The Verge’s report on Anthropic’s office suite, while Sheets and Whiteboards are initially labeled as beta products. A beta is a version that is not final and may still change in features or reliability.
Anthropic is also introducing Interactive Connectors, meaning connections that open partner apps directly inside Claude. The named partners include Adobe, Atlassian, HubSpot, and Salesforce. The practical benefit is that you may need to copy less material or switch between fewer windows. Anthropic also says its tool marketplace has more than 9,000 plugin options, while the new connectors are designed to integrate those services more deeply than earlier extensions.
This moves Claude closer to an office suite without merely recreating a traditional word processor. Existing tools such as Claude Docs, Slides, Design, and Motion are being joined by spreadsheets, meetings, and data analysis. The Verge gives another example: turning a presentation into a one-minute explainer video. The supplied reports do not independently verify how reliably these multi-step processes work outside prepared demonstrations.
Meetings are processed while people speak
Claude Meetings adds Artificial Intelligence (AI) to a video call as a visible participant. According to the report on the Claude Meetings launch, Claude follows the discussion and starts working before the call ends. It can draft the plan being discussed, lay out a timeline on a whiteboard, answer questions in the meeting chat, and retrieve information such as last year’s figures from connected work apps.
After the call, notes, action items, and generated materials are available in claude.ai. One concrete use case is a project meeting in which the team discusses dates and responsibilities: Claude can build the timeline and record the next steps at the same time. During a brainstorming session, the whiteboard can develop as people talk and later be converted into a document or slide deck. Both examples come from the functions Anthropic describes, and their usefulness still depends on Claude capturing statements and responsibilities correctly.
Anthropic has included some visible limits. Claude identifies itself as a meeting participant and posts a notice in the chat. It does not join calls with external guests unless an administrator authorizes it, and it automatically leaves if an external guest joins later. Meetings is launching in beta on Team and Enterprise subscriptions and initially works with Google Meet; support for additional platforms is only described as forthcoming.
For businesses, Anthropic also lists audit logs, which are traceable records of activity, isolated virtual machines, and an interface for compliance controls. Those measures may support oversight and internal reviews, but they do not settle every privacy concern. The sources provide no specific information about Swiss availability, Swiss data locations, or supported languages. Organizations in Switzerland therefore still lack details on how these tools fit internal privacy requirements and confidential discussions.
New models offer more performance at varying costs
Alongside the office tools, Anthropic is introducing Claude Fable 5.2 and Claude Mythos 5.2. Both use the same underlying model, according to an analysis of the model release, but apply different safeguards. Fable blocks high-risk cybersecurity requests and much biological research. Mythos removes some of those restrictions and is available only through verification programs for cybersecurity and life sciences.
The models are intended for long technical, research, and agentic tasks. Agentic work means that software carries out several steps with tools with a degree of independence. MarkTechPost reports a one-million-token context window and a maximum output of 128,000 tokens; tokens are small units of text used to measure input, output, and billing. The release also reports strong results across several evaluations, although the comparison scores were published by Anthropic itself.
Fable 5.2 reportedly reaches 49.1 percent on FrontierCode 2.0 and 41.3 percent on the Remote Labor Index. It does not lead every evaluation: Gemini 4 Carbon scores 53.9 percent on AutomationBench versus Fable’s 51.4 percent, while Opus 5.5 has a slightly higher AA-Briefcase rating. Such measurements help compare models, but they do not fully represent a typical team meeting or your particular spreadsheet.
The sources also present a mixed cost picture. According to the assessment of Fable 5.2 and Mythos 5.2, usage costs $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for the previous models. At maximum quality, however, an average task reportedly costs $26.50, or about 2.6 times as much as with Opus 5.5. A lower listed price per token does not necessarily produce a cheaper completed task.
The reports also describe access differently. MarkTechPost says Fable 5.2 runs only through hosted interfaces and names Claude Platform, Amazon Web Services, Google Cloud, and Microsoft Azure. The Decoder says Fable is available to all users, while still describing Opus 5.5 as the preferred model for everyday use. The supplied material therefore does not make it clear whether you can select Fable directly in ordinary Claude subscriptions, or in which ones.
Pros and Cons of the expanded Claude platform
Pros:
- Less app switching – Spreadsheets, whiteboards, data sources, and partner apps can come together in one work environment.
- Faster follow-up – Meeting notes, tasks, and timelines can take shape during the conversation rather than after everyone leaves.
- Shared workspaces – Teams can work with Claude directly on a board or spreadsheet instead of manually transferring results.
- More capable models – Published evaluations show substantial Fable 5.2 gains on long and complex assignments.
Cons:
- Beta status – Meetings, Sheets, and Whiteboards are not yet presented as final, fully proven products.
- Confidential material – Conversations, business figures, and connected apps place particularly sensitive information in one environment.
- Unclear total cost – Token prices, computing effort, and optional speed modes make completed tasks difficult to compare.
- Limited availability – Meetings starts on selected subscriptions and only with Google Meet, while Swiss details are absent.
Practical value depends on control and task selection
If you are just getting started, choose a small task whose result is easy to check. You could ask Claude to organize a nonconfidential brainstorming session on a whiteboard or build a simple spreadsheet with multiple tabs and formulas. Review labels, calculations, dates, and responsibilities manually before using the output with your team.
More advanced users can connect several functions into one workflow. A team could discuss a plan in Google Meet, have Claude create a timeline in parallel, and then turn the finished whiteboard into a slide deck. For data projects, Claude Data can visualize information from connected systems; that makes clear access permissions and checks of the selected data especially relevant.
Fable 5.2’s gains are most visible on larger assignments. In the Remote Labor Index of real freelance projects, the model completed 41.3 percent of assignments at a quality a paying customer would accept. The test reports particular progress in audio and music, architecture and interior design, and 3D modeling. Industrial product design, by contrast, remains among the areas the system can barely handle, according to the report.
A result of 41.3 percent is notable, but it also means that most tested assignments were not completed at that standard. The evaluation shows potential for bounded professional tasks, not the dependable replacement of entire jobs. In daily work, the useful question is therefore less whether Claude tops a ranking and more whether you can inspect intermediate results and catch errors without creating excessive extra work.
The new features make Claude a more credible work platform because meetings, ideas, spreadsheets, and data are no longer treated merely as attachments to a chat. Combining office tools with more capable models could shorten recurring handoffs, but it also creates a central location for sensitive content and possible misinterpretations. The main unresolved issues are the reliability of the beta features, the real total cost, and the specific privacy conditions for users in Switzerland.
Sources
- Claude-App bekommt Tabellen, Whiteboards und eingebettete Partner-Apps – The Decoder, 2026-11-10
- Anthropic launches Claude Meetings to turn workplace discussions into plans and tasks – TechCrunch, 2026-11-10
- Anthropic has its own AI-powered office suite now – The Verge, 2026-11-10
- Anthropic Launches Claude Fable 5.2 and Mythos 5.2: Same Model, Different Safeguards – MarkTechPost, 2026-11-10
- Zurück an die KI-Spitze: Anthropic stellt Claude Fable 5.2 und Mythos 5.2 vor – The Decoder, 2026-11-09
- Anthropics Fable 5.2 erledigt über 40 Prozent echter Freelance-Aufträge in Profiqualität – The Decoder, 2026-11-10
Image: Yan Krukau via Pexels