T7

Keep your source code and prompts on your Mac

None of these candidates is documented to meet both the US$20 monthly budget and the requirement that source code and prompts never leave this device. Privacy Mode, no training and zero retention each control something different, and none of them establishes that nothing is transmitted.

Sources checked 2026-09-23

The task

Edit an existing private repository on a Mac for up to US$20 a month. Sending source code or prompts to any remote machine, including a company-owned server, is prohibited.

Official sources checked 23 September 2026. Budgets assume a US personal account, USD before tax and monthly billing without annual prepayment. Region, eligibility and checkout prices can differ, and nothing here is a generation, quality or workload test.

Candidates

Meets the documented conditions is not the same as a completed task: it means the conditions listed in the official pages are satisfied, and the limits below still apply.

Cursor standard AI workflow

Does not meet this task

This month: Pro is US$20 a month: within budget.

Does not meet the task. Requests pass through Cursor backend even when you supply your own API key, and privacy mode, no-training and zero-data-retention settings carry their own conditions and abuse-review exceptions. Temporary server-side caching is not on-device processing.

  • A data-use statement written on 3 September 2026 is the basis here; we did not capture network traffic.

Read the Cursor listing →

Devin Desktop / Devin Local

Unconfirmed; not recommended yet

This month: Pro is US$20 a month: within budget.

Unconfirmed and not recommended yet. Desktop requires backend connectivity, and while Local means local execution, the current evidence does not establish that inference and prompt construction stay on the Mac. Sharing a conversation uploads a processed transcript.

  • Backend connectivity alone does not prove that source code is sent, and we did not capture traffic to settle it.

Tabnine SaaS / private deployment

Unconfirmed; not recommended yet

This month: Current public purchase routes emphasise annual team plans; currency and a US$20 monthly route were not verified this round.

SaaS does not meet the task because inference transmits code. A private remote cluster still leaves this Mac, and an air-gapped deployment would need the model endpoint, logs and telemetry, Mac support and a quote checked - a qualifying single-device setup is unconfirmed.

  • Self-hosting a server is not the same as running the model on the client device.

Read the Tabnine listing →

The hard boundary

Local file editing is not local inference. No training is not zero retention, and zero retention is not zero transmission. Turning off indexing, telemetry or sharing one feature does not establish that every AI request runs offline, and a policy requirement cannot be satisfied by a setting whose scope is narrower than the policy.

Next check

Reconsider a candidate only when a supported, priced configuration keeps the model, prompt construction, retrieval and execution on the same device. Then verify the features you need with outbound networking blocked and inspect the traffic. We have not run that check, and the requirement should not be relaxed to manufacture a recommendation.

Where every claim comes from

Each link was opened and read on the date shown. The note describes what that page actually supports.

Other guides

← All task guides