Exchange Ranking
Top Exchanges
24h Volume Ranking
Popularity Ranking
Exchange BTC Balance
Proof of Reserves
Decentralized Exchanges
Funding Rate
Funding Heatmap
Liquidation Data
Max Pain
Long/Short Ratio
Whale L/S Ratio
Binance/Okex/Huobi L/S
Bitfinex Margin L/S
ETF Tracker
Solana ETF
XRP ETF
Hong Kong ETF
Bitcoin Treasuries
Crypto Reversal
Ethereum Reserves
HyperLiquid Wallet Analysis
Hyperliquid Whale Watch
Large Transactions
On-chain Movement
Bitcoin ROI
Stablecoin Market Cap
Options Analysis
News
Articles
Economic Calendar
Features
Wallet
Contract Calculator
Security
Collections
Watchlist
FollowingGoogle Cloud launches Gemini agent that supports the Anthropic Claude model
On Thursday, Google Cloud announced the launch of a Gemini agent that runs based on the Anthropic Claude model for Gmail, Docs and Slack. The company said nearly 80% of its customers are using its artificial intelligence products.
Core Highlights
- The agent handles multi-step tasks within Workspace app and is accessible through Slack and Microsoft 365.
- It will choose between Gemini and Claude models based on specific tasks, and plans to support other models in the future.
- Google Cloud has not yet announced pricing or a full release date.
Detailed explanation of the functions of the Gemini agent
CEO Thomas Kurian introduced the product in a company blog post based on his keynote speech at the "Gemini at Work 2026" event. He described a single-day agent capable of answering questions, processing knowledge-based work, creating images and media, and writing and running code. Kurian writes that users only need to provide it with "goals" rather than specific "instructions."
Inside Workspace, the agent can work in Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar. Employees can also interact with it via Slack, the Microsoft 365 suite or the command line. It runs in the cloud, maintains a single memory across devices, and breaks down long-term tasks into temporary sub-agents, who complete the work in hours or even days.
In addition, a separate "colleague model" can assign the agent a dedicated Workspace account, email address, and calendar, but it can only see what the team shares to it. The agent can also connect to external business systems such as Salesforce, ServiceNow and Snowflake. It is reported that the product is still in the private preview stage and plans to be more widely promoted to Workspace customers who choose commercial and enterprise versions of the plan.
Kurian talks about costs and controls
Kurian's promotion focuses on control rights and costs. He pointed out that the success or failure of enterprise-level agent projects depends on two key issues: whether the company can govern them and whether it can afford the costs.
Each agent has an independent identity, and its operations are recorded in the audit trail. When the project reaches the expenditure limit, the budget limit will automatically suspend the operation of the agent. Kurian believes that matching each task to the right model can both improve the accuracy of complex tasks and reduce the cost of simple tasks.
There are still some details missing. The company did not disclose separate pricing, comprehensive release dates or independent benchmark data, and the user adoption data it mentioned described existing AI products rather than the newly released agent.
Competitive situation with Microsoft and OpenAI
The release follows a series of competing products aimed at the same corporate buyers. On September 25, Microsoft added the durable Autopilot agent to the redesigned Copilot; four days later, on October 6, OpenAI launched the always-on Dots agent. The Anthropic model called by the Gemini agent also entered Workspace's Docs, Sheets and Slides applications on the same day.
