OpenAI shipped something on June 18, 2026 that isn't going to get the same attention as a new model release but matters more to...
OpenAI shipped something on June 18, 2026 that isn't going to get the same attention as a new model release but matters more to most enterprise teams right now: credit usage analytics and updated spend controls for ChatGPT Enterprise. Available immediately through the Global Admin Console, these tools give admins a real breakdown of how ChatGPT and Codex credits are being consumed across their organization by individual user, by product, and by model.
If you've been managing a ChatGPT Enterprise deployment by watching invoices arrive and trying to work backwards from the total, this is the update you've been waiting for. If you haven't been paying close attention to per-user AI spend, this is the update that will make you start.
What the Global Admin Console Now Shows You
The analytics side of this release lives in the Global Admin Console. Admins get a unified view of credit consumption across ChatGPT and Codex not just totals, but breakdowns by individual user, product, and AI model. You can track usage trends over time, see who your top consumers are, and spot emerging patterns before they become budget problems.
OpenAI is also exposing the same credit usage data through a unified Cost API, which means teams that want to pull this into their own dashboards, finance systems, or internal reporting tools can do that without manually exporting from the console. That's the right call enterprise finance teams don't want another portal to check. They want the data where they already work.
Zipline, one of OpenAI's enterprise customers, noted their engineering team has been using Codex heavily since January. That kind of sustained, organization-wide usage is exactly the scenario where not having usage visibility creates real problems you don't know which teams are driving value and which are running up costs on low-impact tasks.
How the Spend Controls Actually Work
The controls are more granular than the previous version. Admins can now set a default credit limit for the entire workspace a baseline that applies across the organization. On top of that, they can configure separate limits for specific groups, so the engineering team can have a higher ceiling than, say, the marketing department if the workloads justify it. And individual overrides are available for power users who legitimately need more capacity without triggering a blanket limit increase for their whole department.
The employee-facing side is worth noting too. Individual users can now see their own credit usage against their allocated budget inside workspace settings. They can also request additional credits directly, with the ability to add context about what they're working on. That's a better flow than the previous situation, where a user hitting a limit would need to go to IT or their manager with no clear process for what happened next.
The three-tier structure workspace default, group limits, individual overrides is the right architecture for organizations with heterogeneous AI usage patterns. Most enterprises don't have uniform AI consumption across departments, and a one-size-fits-all credit limit either over-restricts the teams that need access most or under-restricts the teams that are generating cost without proportionate value.
Why This Release Matters More Than It Looks
The timing isn't accidental. OpenAI is heading toward a public listing, and the story it needs to tell investors is about predictable, scalable enterprise revenue not just impressive model benchmarks. Features like these are what make enterprise customers renew contracts and expand deployments rather than pulling back when the first big invoice arrives.
There's also a competitive dimension. Microsoft, through Azure OpenAI, has offered cost monitoring and budget alerts for a while now. Google's Vertex AI has similar tooling. The fact that OpenAI's own enterprise product is now catching up to the governance and observability standards that cloud infrastructure providers have offered for years signals that the ChatGPT Enterprise platform is being taken seriously as a long-term operational tool, not just a capability showcase.
For enterprise decision-makers, the practical question is straightforward: can you now answer "what are we spending on AI and is it worth it?" with actual data rather than estimates? If the Global Admin Console and Cost API deliver on what's described, the answer is yes and that's a meaningful shift for organizations trying to move AI from pilot to permanent infrastructure.
What to Actually Do With This
If you're a ChatGPT Enterprise admin, the features are live now no waiting for a rollout. First move is to pull the usage breakdown and see if what you find matches your intuition about where AI is being used in your organization. It usually doesn't, and the gaps are informative.
Second, set the workspace-level default before you do anything else. Even a generous limit is better than no limit it creates a signal when something unusual is happening. Then layer in group limits for the teams where you have enough usage data to know what normal looks like.
The Cost API integration is worth prioritizing if your finance or operations team already has tooling for tracking cloud spend. Getting AI credits into the same reporting layer as AWS or Azure costs makes the conversation about AI ROI significantly easier it stops being a separate discussion and becomes part of the normal infrastructure cost review.
Frequently Asked Questions
Q: What did OpenAI announce for ChatGPT Enterprise on June 18, 2026?
OpenAI launched credit usage analytics and updated spend controls for ChatGPT Enterprise. Through the Global Admin Console, admins can view breakdowns of ChatGPT and Codex credit consumption by individual user, product, and AI model. They can set workspace-wide credit limits, configure group-level limits, and create individual overrides. Employees can check their own usage and request additional credits. The same data is available via a unified Cost API.
Q: What is the OpenAI Global Admin Console?
The Global Admin Console is OpenAI's enterprise dashboard for ChatGPT Enterprise customers. As of June 18, 2026, it provides a unified view of credit consumption across ChatGPT and Codex, with breakdowns by user, product, and model, usage trend tracking over time, and tools to set and manage credit limits at the workspace, group, and individual level.
Q: How do the new OpenAI spend controls work?
The controls work in three tiers: a default workspace-wide credit limit, group-specific limits for different teams, and individual overrides for power users. Employees can view their own usage in workspace settings and request additional credits with context about their work.
Q: Is there an API for OpenAI enterprise usage data?
Yes. OpenAI's unified Cost API gives enterprise customers programmatic access to the same credit usage data in the Global Admin Console, making it possible to integrate OpenAI spend data into existing finance tools, internal dashboards, or cloud cost monitoring systems.
References
- OpenAI. New usage analytics and updated spend controls for enterprises. June 18, 2026. openai.com
