OpenAI is preparing to test a single interface that merges ChatGPT Chat with its Work agent, according to an email seen by BleepingComputer. The idea is simple: users should be able to ask a quick question, then move into a longer project without jumping between tools. The company plans to start with a small group of users. Codex will remain separate.
A single chat for quick and long tasks
Chat has been the place for fast answers and brainstorming. Work is aimed at heavier jobs such as research, spreadsheets, reports, presentations and published Sites. The merged version would let users upload files, connect plugins and watch the agent work from the same conversation. That sounds convenient. It also makes usage harder to predict, since a long agent run may eat more allowance than a quick chat.
Costs move into focus
Anthropic has already moved in this direction. It began rolling out a unified Claude experience for Pro, Max and Ultra subscribers, with enterprise access following and Team and Free plans still to come. It also introduced Docs and Slides. Both companies want users to finish tasks inside one product. The difference may come down to price clarity. OpenAI already shares an allowance structure across Work and Codex, and users can buy credits when limits run out. On the Pro $500 plan, Ultrafast uses included usage at eight times the standard rate, then draws credits at six times the rate. In December, Plus, Pro and Business subscribers will share a token-based allowance across ChatGPT and Codex. Instant chat will not count, and usage details will appear under Settings and Usage. OpenAI doubled Work and Codex allowances for some Pro accounts in late October. Even so, some Codex users have complained that allowances drained faster than expected. OpenAI executive Thibault Sottiaux disputed claims of undisclosed cuts. One analysis first called the claims supported but unproven, then shifted to partly confirmed after cache-efficiency problems surfaced.
Adoption lags spending forecasts
The market is growing quickly. Gartner expects spending on AI agents and assistants to rise from $29.2 billion in 2026 to $65.5 billion in 2027. Total AI spending could reach $2.67 trillion this year, up 49.5%. Adoption is less even. An IDC study of 1,900 organizations found only 3.1% at the highest AI-maturity stage, while 61.3% were in the two earliest stages. For now, predictable costs may matter as much as agent capability. If users cannot tell what a task will consume, the merged interface may feel less like a shortcut and more like a budget question. I think that is the test OpenAI has to pass.
