Close vs Twenty for AI agents
As of 11 Oct 2026, Twenty gets an AI agent further: L4 · Runs on its own against L1 · Needs a person.
- Find: Docs for agents: llms.txt, an OpenAPI and an MCP server.
- Access: Can't sign up.
- Use: Work not reached.
- Pay: Payment not reached.
- Leave: Leaving not reached.
Stopped at sign up: blocked. Next: let an agent sign up on its own
A person is needed before any agent work starts: L1 · Needs a person. Our agent's browser signup at app.close.com stopped after the password step, refused by an invisible reCAPTCHA with no challenge to solve: "We couldn't verify this signup." Every later checkpoint, from key to first call, went untested because no account existed. The product's agent surface otherwise holds up well: llms.txt, an OpenAPI spec, Markdown docs, and an official MCP server whose OAuth client registration worked headlessly, though authorizing still needs a signed-in person. The trial needs no card, and paying or leaving were never reached.
Driven 11 Oct 2026 · The drive and its evidence- Find: Docs for agents: llms.txt, an OpenAPI and an MCP server.
- Access: Sign-up in a browser.
- Use: 80% of the work.
- Pay: Payment not reached.
- Leave: Leaves on its own: Cancels on its own, deletes its account.
Got in; payment not reached. Next: pay without a person
Runs on its own: L4 · Runs on its own. Signup, onboarding and the first key need a browser, and invisible reCAPTCHA let our agent through with no challenge shown. Paying is where it stopped: a trial starts without a card, but clicking add credit card at checkout did nothing in our agent's browser, so a person has to add the card. Short of that, Twenty worked headless: REST and an official MCP server let it create a company and read it back, with tools covering records, custom objects, views and workflows. Cancelling and deleting the workspace both worked on their own, with no card ever added.
Gets furtherDriven 2 Oct 2026 · The drive and its evidenceThe two summaries: written and checked by our agents from each product's drive. How we write.
Where each one stops.
CloseStopped at sign up: blocked
TwentyGot in; payment not reached
| Checkpoint | Close | Twenty |
|---|---|---|
| Docs | Headless | Headless |
| Sign up | Blocked | In a browser |
| Verify | Not reached | In a browser |
| Key | Not reached | In a browser |
| First call | Not reached | Headless |
| First action | Not reached | Headless |
| Pay | Not reached | Not reached |
| Cancel | Not reached | In a browser |
| Delete | Not reached | In a browser |
Headless or in a browser.
How our agent passed each checkpoint, and what each product offers an agent.
Checkpoints by route, API, MCP, CLI and SDK
| Close | Twenty | |
|---|---|---|
| Headless (checkpoints) | 1 | 3 |
| In a browser (checkpoints) | 0 | 5 |
| Needs a human (checkpoints) | 0 | 0 |
| Blocked (checkpoints) | 1 | 0 |
| Not tested (checkpoints) | 7 | 1 |
| Work done | Not measured | 80% of the core |
| API | Yes | Yes |
| OpenAPI | Yes | Yes |
| MCP server | Yes (official) | Yes (official) |
| CLI | No | Yes |
| SDK | No | No |
| llms.txt | Yes | Yes |
Bot checks, and where a person was needed.
| Close | Twenty | |
|---|---|---|
| Bot check | reCAPTCHA (invisible) stopped it at sign up, without a visible challenge | reCAPTCHA (invisible) let it through at sign up |
| Wall type | A check of our network and browser | No bot-check wall |
| First wall | Invisible reCAPTCHA refuses the sign-up | Got in; payment not reached |
| A person helped | No | No |
A check of our network and browser: It judged where our agent connects from and its browser, not a person. An agent that works on its own from a data center doesn't get through unless the owner lets signed agents in. How agents in a person's own browser fare isn't measured yet. Where we drive from.
Level and part scores.
| Close | Twenty | |
|---|---|---|
| Level | L1 · Needs a person | L4 · Runs on its own |
| Agent-ready setup | 90 of 100, good | 100 of 100, good |
| Standards | MCP client registration (behind) · OpenAPI 3.1.0 (1 version behind) | None recorded |
Holds up: not measured in these drives. Docs match is measured in an Audit. What the scores mean.
Pay, cancel and delete.
| Close | Twenty | |
|---|---|---|
| Pay | Payment not reached | Payment not reached |
| Free plan or trial | Trial, no card | Trial, no card |
| Cancel | Not reached | In a browser |
| Delete the account | Not reached | In a browser |
How each one takes payment
- Close: Payment not reached.
- Twenty: Payment not reached.
Close vs Twenty: the short answers.
Which can an AI agent use, Close or Twenty?
Twenty gets an AI agent further: L4 · Runs on its own against L1 · Needs a person. With Close our agent was stopped at sign up (blocked); with Twenty it got in (payment not reached). Driven 11 Oct 2026 (Close) and 2 Oct 2026 (Twenty).
Can an AI agent sign up for Close and Twenty on its own?
Close: no, blocked. Twenty: yes, in a browser.
Do Close and Twenty have an MCP server?
Close: yes (official). Twenty: yes (official).
Can an AI agent delete its Close or Twenty account?
Close: not reached. Twenty: yes, in a browser.
The drives and the category.
Put your own shortlist side by side.
A Stack check drives the tools you're choosing between and shows which ones your agents can use. $490 for three tools, ready in 24 hours.