Opportunities › Privacy-respecting AI code completion platform › Proposed app
Plainkey
Code completion that never trains on your code, with a receipt per request
- Score
- 37
- confidence 50
- Evidence
- 3
- items on the page
- Rivals
- 2
- named in sources
- First version
- 35 d
- estimate
- Type
- Desktop app
- PWA: runs in the browser, installs on Mac and Windows
- Market since
- 2021
- Emerging market · AI estimate
ScreensConcept
Desktop app: Code completion is used during development work at an editor; developers spend long stretches writing and reviewing code at a desk. Sample screens on a computer, in the proposed brand. Their numbers are sample data.
Receipts
Nothing stored0 kept
About this appAI proposal
Plainkey is a code completion endpoint you plug into the editor you already use. It runs open-weight code models and promises one thing in writing: your code and chats are never used for training. There is no opt-out because nothing is collected for training.
A phone-friendly dashboard shows a receipt for every request: what was sent, which model answered, and when it was deleted. Retention is off by default. If you want history for debugging, you turn it on yourself, with a plain switch and an expiry date.
It is for developers and small teams stuck between tools with unclear data terms and the work of self-hosting. Consent screens use equal buttons and plain sentences. Export and delete take one tap each.
- Tagline
- Complete your code. Keep your code.
- Voice
- PlainDirectCalmHonest
- Why this name
- Plain says what it is: no tricks. Key points to ownership and to the API key you connect with. Two short, pronounceable words joined.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
No training, in the terms
Your code and conversations are never used to train any model. The promise is a short page you can read in a minute, not a buried clause.
No member yet
A receipt for every request
See size, model, retention state and deletion time for each completion. Receipts are signed so a team can keep them as a record.
No member yet
Honest consent screens
Every choice has equal buttons and plain wording. Nothing is pre-ticked, and declining never hides features.
No member yet
Retention you control
Default is zero retention. Opt in per project for a set number of days, then it deletes itself.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#2F5D50Pine · primary#F6F1E7Paper · surface#1F2A24Ink · text#D98E04Amber · accent#1E6F8CLake · data#B3412EBrick · warning
Typography
Aa
Plainkey
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test, not based on any known price.
Free
Free
- Limited completions per day
- Zero retention
- Receipts and export
Solo
$12 a month
- Higher daily limits
- Faster models
- Optional retention per project
- Never-send rules
Team
$24 a month
- Per seat, shared policies
- Signed receipt archive
- Admin consent controls
- Priority inference
Would you pay Solo · $12 a month?
Your answer tells PWAS which prices people accept. It changes no score.
No answers yet.
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$194,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Partners and integrations. It works through editor extensions that accept a custom endpoint. Listings in their guides, plus a one-page setup guide per editor, put it in front of developers already looking for private options.
- Do they become customers?
- Those reached are developers who set it up themselves, so they can pay directly. Test whether free users hit limits and upgrade, and whether teams adopt shared policies.
- First customers
- Freelance developers and small agencies with client code under NDA, using an editor that supports a custom completion endpoint.
- Where to reach them
- Hacker News: Post a Show HN with the plain terms page and a sample receipt, and answer questions on data handling.
- Editor extension communities: Publish setup guides in forums and issue trackers where people ask for private completion.
- Developer newsletters: Offer a short write-up on how zero-retention completion works, with a free trial key.
- First milestone
- Twenty free users connect an editor, and five freelancers or small agencies pay for Solo or Team within 60 days.
Why people already pay for this
- Public claimCodex uses dark UI patterns and unclear consent mechanisms for data collection
- Public claimUsers are being silently trained on by code completion tools with unclear disclosure
- Public claimData collection practices in existing tools are driving developers to consider self-hosted alternatives
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · Model inferenceHosted GPU inference for open-weight code modelsCompletions need low latency and a model we can promise is never trained on user data.
- This app · Streaming gatewayEdge proxy with OpenAI-compatible APIEditors connect through a custom URL, and requests must stream and be dropped after use.
RisksAI inference
- TechnicalCompletion quality and latency must match what developers already use, or privacy alone will not keep them. Open-weight code models may lag at first.
- CommercialGPU inference costs may exceed subscription revenue at low prices, especially on the free tier.
- CommercialDevelopers may prefer free self-hosting, so paying for a hosted service must be worth the saved effort.
- PlatformSome editors or extensions may not allow custom endpoints, which limits where Plainkey works.
- RegulatoryA no-training promise is a legal commitment. Logs, abuse handling and subprocessors must match it exactly, and privacy law applies.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.