A code editor that is also an agent harness, so it sits on the boundary between the editor a person works in and the harness that holds an agent's loop.
Our description, from the vendor's documentation, as of 2026-09-03.
Runs onThe places the vendor's own pages say the product runs: your machine, a container environment you run, or a cloud service the vendor runs. It is not a limit.
Models runWhere the vendor's own pages say the product's models run, not a limit: your machine, a container environment you run, the model company's cloud (servers of the company that makes or serves the model), or a cloud provider's model service. For a cloud, read the data and compliance lines of the company running it, linked by its name.
Several providers in one sessionOne session can use models from different providers, for example one agent on one maker's model and another on a second maker's. Switching the single model for the whole session does not count on its own. A tool that is itself a gateway or library routing to the providers its page names counts.
Several endpoints built inThe tool itself can hold more than one provider or address, so nothing has to sit in front of it to reach different providers.
Hold several endpoints in the product: How to set it up(cursor.com page, read 2026-10-22) Settings the guide names: Cursor Settings, Models.
Compared with similar tools
Same role (harness / agent client) at the same layer (Coding agents and assistants). Up to 4 others, in alphabetical order; nothing is ranked.
Product
Several providersOne session can use models from different providers, for example one agent on one maker's model and another on a second maker's. Switching the single model for the whole session does not count on its own. A tool that is itself a gateway or library routing to the providers its page names counts.
Several endpointsThe tool itself can hold more than one provider or address, so nothing has to sit in front of it to reach different providers.
Model on your machineThe tool can use a model running on your own computer, for example through Ollama, LM Studio, oMLX or llama.cpp.
Local and cloudA local model (on your machine or in a container environment you run) and a cloud model (in its model company's cloud or a cloud provider's model service) can both be used in the same session or setup.
Words in the cells: YesA page we read says so., PartlyA page says part of it, for example separate models per agent without saying the providers can differ., Via a proxyOnly with a proxy, or a local server that imitates a cloud provider's interface, in front of the tool., NoA page we read says the product does not support it., ClaimedA vendor page says so on a marketing or overview page, and we found no guide showing how to set it up., Not foundWe read pages for it and found nothing there. It says nothing about whether the product has the feature., n/aThe question does not apply to the product, for example because it is a model server and not a tool that connects to models., BlankNothing is published for it yet.
Data and compliance
Compliance: SOC 2 Type II (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.); ISO/IEC 27001:2022 (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.); ISO/IEC 42001:2023 (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.); HIPAA BAA (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.; Enterprise; under a conditionThe mark holds only under a setting or an agreement the line names.); DPA (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.).
Data: Training use only under a conditionEvery sentence found holds only under a condition, such as a setting or a signed agreement.; default retention not stated on the pages readWe read a page that could answer and found no sentence that does. It says nothing about what the maker does.; zero-retention option only on some plans or optionsNothing found covers the product as a whole; what was found holds only for the plans, options or regions named. (Custom API keys (BYOK)); data residency stated on the maker's pageA sentence on the maker's page answers it for this product..
Model and where it runs: Installed: an IDE; model provider stated for the whole companyThe only sentence found is about the maker as a whole and names no product, so check on the page whether it covers this one.; where it can run only on some plans or optionsNothing found covers the product as a whole; what was found holds only for the plans, options or regions named. (Custom API keys (BYOK), “Self-Hosted Machines”).
Details: What the maker's own pages state, not a compliance determination, unverified after 2026-12-23: trust page, cursor.com.
What the maker's own pages said on the dates shown, not a compliance determination.
Also covers
The vendor's own pages present these as parts or modes of this product, so each is described here and has no page of its own.
Cursor Cloud Agents
Cursor's hosted, asynchronous mode, separate from the Cursor editor: a task is handed to it and left running in a cloud VM rather than on the local machine.
Our description, from the vendor's documentation, as of 2026-09-23.
Compliance: For Cursor, SOC 2 Type II (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.); ISO/IEC 27001:2022 (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.); ISO/IEC 42001:2023 (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.); HIPAA BAA (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.; Enterprise; under a conditionThe mark holds only under a setting or an agreement the line names.); DPA (company-wideThe mark rests on a sentence that names no product, so check on the page whether it covers this one.).
Data: For Cursor, training use only under a conditionEvery sentence found holds only under a condition, such as a setting or a signed agreement.; default retention not stated on the pages readWe read a page that could answer and found no sentence that does. It says nothing about what the maker does.; zero-retention option only on some plans or optionsNothing found covers the product as a whole; what was found holds only for the plans, options or regions named. (Custom API keys (BYOK)); data residency stated on the maker's pageA sentence on the maker's page answers it for this product..
Model and where it runs: Hosted by the maker; for Cursor, model provider stated for the whole companyThe only sentence found is about the maker as a whole and names no product, so check on the page whether it covers this one.; where it can run only on some plans or optionsNothing found covers the product as a whole; what was found holds only for the plans, options or regions named. (Custom API keys (BYOK), “Self-Hosted Machines”).
Details: What the maker's own pages state, not a compliance determination, unverified after 2026-12-23: trust page, cursor.com.
What the maker's own pages said on the dates shown, not a compliance determination.
Cursor lets users add their own API keys for the providers its page lists (OpenAI, Anthropic, Google, Azure OpenAI and AWS Bedrock), and those providers' models then appear in the model picker. (cursor.com page, read 2026-09-24)
What you set up
Cursor's API keys page says that once a key is added for one of the providers it lists (OpenAI, Anthropic, Google, Azure OpenAI or AWS Bedrock), Cursor applies that key to models from that provider; the page narrows this elsewhere: OpenAI to standard, non-reasoning chat models, and custom keys to chat models only. (cursor.com page, read 2026-09-24)
Models from different providers in one session
Cursor's subagents page advises setting a subagent's model to inherit when it should reason as strongly as its parent, and to a specific model ID when it needs that model's capabilities whichever model the parent uses. (cursor.com page, read 2026-10-21)
On Cursor's subagents page, a subagent's model setting can be inherit or a specific model ID, which Cursor then uses for that subagent; the page's examples are composer-2 and gpt-5.6-sol. (cursor.com page, read 2026-10-21)
Cursor's models page lists GPT-5.6 Sol under OpenAI and Composer 2.5 under Cursor. (cursor.com page, read 2026-10-21)
Plan, policy and billing
Cursor's model and integration management page says team settings, Enterprise only, let an admin turn models on or off, choose defaults for newly added providers and models, and decide whether members may bring personal API keys (BYOK). (cursor.com page, read 2026-10-23)
Cursor's model and integration management page says Enterprise teams can block their members' use of personal API keys with outside providers, naming OpenAI, Anthropic, Azure and AWS Bedrock. (cursor.com page, read 2026-10-23)
On Teams and Enterprise plans, a Cursor Token Rate applies to third-party model requests, at $0.25 per million tokens. (cursor.com page, read 2026-10-19)
About the product
Cursor's documentation home invites the reader to get Cursor for their own computer. (cursor.com page, read 2026-10-24)
Cursor's cloud agents page says cloud agents run in isolated virtual machines in the cloud rather than on the user's local machine, each with a complete development environment. (cursor.com page, read 2026-10-24)
Our reading of vendors' pages, as of the dates shown. Confirm on the vendor's page before relying on any line that cites it. To report an error write to report@crinaro.ai