Cheaper small models, chat tools and agent wallets arrive

Small models get cheaper

Anthropic has released Claude Haiku 5.5, its fastest and most affordable small model, built for high-volume tasks like data queries and customer support. On average the model costs about 75 percent less than Haiku 4.5, and for requests with prompts up to 100,000 tokens, prices drop by up to 90 percent. Anthropic points out that the model uses an updated tokenizer that consumes slightly more tokens per task, and The Decoder notes that real-world savings are likely smaller than the per-token prices suggest.

The chat window becomes a tool

Meanwhile OpenAI is rolling out GPT-6 in ChatGPT with a feature called Intelligent UI, which presents answers as graphics, buttons, charts and forms instead of plain text. Users can also have small tools built right inside the chat, like a savings calculator, and GPT-6 can also respond while still "thinking." For those of us who talk with customers every day, a fair question is what they will expect to see when they ask a chatbot about a loan.

Agents get their own wallets

At the same time, First Digital, the Hong Kong-based stablecoin issuer, says it is ready for its stock market debut. Its listing will happen through a special purpose acquisition company (SPAC) merger with CSLM Digital Asset Acquisition Corp III, and the Nasdaq listing is designed to help fund the expansion of Finance District, its ecosystem for the agentic economy. Four Finance District products are already live, including Agent Wallet, a multi-chain wallet that AI agents operate via the Model Context Protocol.

What to watch in your own shop

The price of a model, the screen a customer sees and the wallet an agent might hold are separate layers of the same stack. A good exercise for this week is to list the high-volume jobs in your team and ask who would sign off if a cheaper model took one of them over.