OpenAI Models in Cursor: What Happens Nov 12

OpenAI told Cursor users on 29 August 2026 that it plans to stop supplying its models to the editor, and it put a date on it: 12 November 2026. The wording matters more than the headline does. OpenAI models not available in Cursor is a scheduled outcome with a proposed date attached, not something breaking in your editor this morning; until that date arrives, every GPT model you already use in Cursor keeps working the way it worked last week. What changes afterwards is narrower than the coverage suggests, and the replacement routes OpenAI documents cover far less of Cursor than you would guess from the announcement alone.

Key Takeaways:

• OpenAI has proposed 12 November 2026. Its own help page adds that Cursor "may choose to end access sooner", and that a confirmed termination date gets published only once both companies agree on one.
• Cursor's model catalogue mostly survives. Anthropic, Google, Moonshot and Z.ai models stay, as do Cursor's own Composer and the Grok line. Cursor's CEO puts OpenAI at roughly 5% of user traffic.
• The documented workarounds apply to local Chat and Agent only. Tab, Auto, Cloud and Background Agents, Automations, the CLI and Cursor's API keep using models Cursor supplies, and cannot use your credentials at all.
• Switching to your own OpenAI account moves the invoice with it. A ChatGPT subscription does not include API usage.
• As of 30 August 2026 Cursor's changelog had said nothing about any of this. Its newest entry is dated 27 August.

What OpenAI Actually Put in Writing

The help centre article is short and unusually blunt for a vendor notice: "OpenAI is planning to wind down our contract providing OpenAI models to Cursor. We have proposed a transition period of November 12, 2026. During that period, we would continue providing Cursor access to the OpenAI models it uses today. Cursor may choose to end access sooner. We will share the official termination date as soon as it is confirmed between both companies."

Read that twice, because three separate things in it are conditional. The date is a proposal rather than a deadline. The cut could land earlier, at Cursor's discretion and not OpenAI's. And no confirmed termination date exists yet from either company. Anyone telling you GPT access in Cursor ends on 12 November is repeating a proposal as though it were settled.

OpenAI posted the same summary to its own account that night. Ownership triggered it: SpaceX completed its acquisition of Cursor's parent company in mid-August, and OpenAI invoked a change-of-control provision, saying on its own site that it could not be confident the models would be used within its terms of service. Cursor's CEO Michael Truell replied publicly a little over an hour later, saying his team was "speaking with the OpenAI team to resolve this" and that OpenAI models serve about 5% of Cursor user traffic. Two weeks passed between the acquisition closing and the notice going out.

What Survives in Cursor, and What Goes

The catalogue does not collapse. Cursor's docs still list Anthropic, Google, Moonshot and Z.ai models next to Composer and the Grok line that arrived with the new owner, and none of those are affected by an OpenAI contract.

The pricing page is a useful tell. The $20 Individual plan advertises "Generous limits for Grok", "Grok Bot access" and "Access to frontier models", and never names an OpenAI model once. OpenAI's own logo, meanwhile, still sits in the customer strip at the bottom of the page. That page was written before this week and nobody has been back to it since.

One more detail worth catching while you are there: GPT-5.6 Sol carries promotional pricing in Cursor's model docs through 21 November 2026, which is nine days after OpenAI proposes to stop supplying it. Vendor pages get maintained by different teams on different weeks. A date you read in a doc is worth re-checking before you plan a quarter around it.

The Three Routes OpenAI Documents

OpenAI names three ways to keep using its models inside Cursor after the wind-down, and prefaces all of them with a warning that the guidance is "directional" and Cursor may change the functionality over time.

The first is your own OpenAI API key. You create a key on an API account with billing configured, open Cursor Settings, go to Models, paste the key, enable the model you want, then select it in Chat or Agent. Available models and rate limits follow your OpenAI organisation rather than your Cursor plan.

The second is the Codex IDE extension, which installs inside Cursor and runs its own agent panel. It authenticates with either an eligible ChatGPT subscription or an API key, and it deliberately sidesteps Cursor's model picker: OpenAI is explicit that it "does not change the model used by Cursor Chat, Agent, Tab, Auto, Cloud Agents, or other Cursor services". You get Codex next to Cursor, not Codex inside Cursor's own features.

The third is a gateway. If your organisation already routes model access through Amazon Bedrock, Azure, or another OpenAI-compatible endpoint, Cursor can point at a custom base URL. OpenRouter and Vercel AI Gateway may work, with the caveat that some model-specific controls do not survive the trip through a custom endpoint.

Where Your Own Credentials Stop Reaching

This is the part the announcement coverage mostly skipped, and it is the reason a personal API key is not the clean substitution it sounds like.

Both key-based routes are limited to local requests. In OpenAI's phrasing, Cursor "doesn't support 'Bring your own key' in Cursor Tab, Auto, Cloud or Background Agents, Automations, the Cursor CLI, or Cursor's API and SDK", because those features use models supplied or routed by Cursor and cannot accept your credentials. Gateway setups inherit the identical restriction.

So the practical loss is not GPT in the chat sidebar; you can buy that back with a key and a credit card. The loss is GPT in everything that runs when you are not sitting there. Background agents, scheduled automations, terminal sessions, anything driven through Cursor's API: all of it keeps whatever models Cursor decides to route, and no configuration on your side changes that.

Which is a real gap if the work you cared about was the unattended kind. Something has to hold a machine, a model and a schedule together while you are asleep, and an editor that routes its own background jobs is not built to hand you that steering wheel. That is the specific hole Zentor fills: a hosted cloud AI computer where the always-on work runs on infrastructure you configure, not on a routing policy shipped inside somebody's IDE. It sits beside your editor rather than replacing it, and it does not care which company signed what contract with whom.

The Invoice Moves With You

Pasting a key changes who bills you, and that catches people out. A ChatGPT subscription does not include OpenAI API usage; those are separate products on separate meters. Your Cursor subscription keeps charging for Cursor, and OpenAI starts charging you at API list prices for the requests that used to be bundled.

For light chat use that is often cheaper. For heavy agent loops it is usually not, because the requests that burn tokens fastest are the long agentic runs, and those are exactly the ones a per-token bill makes visible. Nobody can tell you which side you land on without your own usage numbers, so pull them before November rather than after your first surprise invoice.

There is a second-order effect too. Once the models come from your own account, model availability tracks your OpenAI organisation's access rather than Cursor's negotiated catalogue. If your org does not have access to a given model tier, it will not appear, and that is a support conversation with OpenAI instead of with Cursor. Splitting one bill into two also splits the person you escalate to, which is the sort of thing that only becomes obvious at 2am. Keeping execution on a machine you control, with Zentor as the layer that owns the runtime rather than a vendor's plan tier, keeps that escalation path in one place.

Cursor's Own Channels Have Not Caught Up

If you went looking for this on Cursor's site, you would not have found it. Search for OpenAI models not available in Cursor and what comes back is press coverage of a corporate dispute, not a settings page telling you what to change.

The newest changelog entry on 30 August is dated 27 August and covers Cloud Agents starting without a repo. The models documentation still lists GPT-5, GPT-5.4 and GPT-5.6 variants with no annotation. Nothing on cursor.com acknowledges the transition period at all, which is consistent with a company that says it is still negotiating rather than announcing.

Treat that silence as information, not reassurance. Truell's post said the two teams are talking, so the terms may move again. Screenshot the pages you plan around; we did, because vendor pages get rewritten and old wording disappears.

The broader pattern is not new. We wrote about Claude silently switching models on Opus 5 fallback and about what a tokenizer change did to Sonnet 5 migration costs, and both had the same root shape: the model behind a product is a supplier relationship, and suppliers change. If you want Codex specifically, our notes on running Codex agent workflows cover the extension path in more depth.

Frequently Asked Questions

Is GPT gone from Cursor right now?

No. OpenAI has proposed 12 November 2026 as the end of the transition period, and says it will keep supplying the models Cursor uses today until then. Nothing has been switched off as of 30 August 2026.

Why is OpenAI cutting Cursor off?

SpaceX acquired Cursor's parent company, and OpenAI invoked a change-of-control provision in the contract. Its public statement cites its experience with Musk-owned companies and says it cannot be confident SpaceX will use the technology within OpenAI's terms of service.

Does this affect Claude, Gemini or Grok in Cursor?

No. Only OpenAI's direct supply is ending. Anthropic, Google, Moonshot and Z.ai models remain, and so do Cursor's own Composer and Grok models.

Can I just paste my own OpenAI API key and carry on?

For local Chat and Agent, yes. Not for Tab, Auto, Cloud or Background Agents, Automations, the Cursor CLI or Cursor's API and SDK, all of which use models Cursor supplies and cannot accept your key.

Will my Cursor subscription get cheaper?

Cursor has not said anything about pricing. As of 30 August the plans are unchanged at Hobby free, Individual $20 per month, Teams $40 per user per month.

What to Settle Before 12 November

Two questions decide how much of this touches you. First, what fraction of your OpenAI usage in Cursor happens in local Chat and Agent rather than in background or automated runs? That fraction is the part a key can rescue. Second, is the work that runs unattended important enough that losing OpenAI models there needs a different home entirely?

If the answer to the second is yes, sort it now rather than in November, while you can test a migration against a working system instead of a broken one. Give the automated work its own always-on environment, keep the editor for editing, and stop treating a model picker as though it were infrastructure. Zentor exists for that first half, and it runs alongside Cursor rather than asking you to leave it.

And keep watching the date. OpenAI models not available in Cursor is currently a proposal with a soft edge on both sides: Cursor can end it early, the two companies can still confirm something different, and the only page that will tell you is one neither of them has updated yet.