The building blocks for agent-native economies are increasingly taking shape.
In the last month, there were major advancements in agent verification, decentralized inference, autonomous coding, AI-native financial intelligence, and agent-driven token economies, highlighting the growing convergence of AI and onchain finance.
Here’s a roundup of updates in the AI space within the last month.
● @BitTorrent
• Expanded its utility through the BTTInferGrid decentralized AI inference launch
● @framer
• Introduced Framer 3.0, an update that lets AI Agents work directly on live canvas
● @vercel
• Introduced Eve, an open-source, filesystem-first framework designed for building, running, and scaling durable AI agents in production
● @useAtelier
• Integrated with @clawpumptech to allow AI agents on Atelier to launch their own tokens
● @Fetch_ai
• Introduced ASI:One Ultra, the deepest-reasoning model in the ASI:One AI family
● @MiniMax_AI
• Introduced MiniMax M3, a multimodal foundation model
● @BAI_AGI
• MiniMax M3 officially went live, with a limited-time free access
● @trythreews
• Introduced a drop-in payment modal for x402 paid endpoints
● @predom_io
• Partnered with Orion Agents for prediction-powered campaigns and onchain agent economy infrastructure
● @AethirCloud
• Launched Aethir Mesh, a proprietary LLM API platform providing access to models like DeepSeek and Kimi directly on Aethir’s decentralized GPU network
● @ritualfnd
• Ritual Chain public testnet went live
● @Zai_org
• Introduced GLM-5.2, optimized for long-horizon tasks, complex systems design, and autonomous coding agents
● @tryquantio
• Introduced Quant AI, a conversational financial intelligence platform designed to simplify market research, portfolio analysis, and multi-asset trading
● @virtuals_io
• Released ERC-8126, an AI agent verification standard that lets agents prove security audits, wallet control, and identity without exposing private data
● @Mastercard
• Launched Agent Pay for Machines, an infrastructure built for AI agents to autonomously transact with each other, at machine speed
—
AI agents are evolving from simple assistants into autonomous economic participants, and Web3 is positioning itself as a foundational layer for the next generation of autonomous AI systems.
