# PHNTM One > PHNTM One is a private desktop AI assistant developed by DeCampLabs. It combines a Raspberry Pi 5 with 8 GB RAM and a 10.1-inch touchscreen for local chat, memory, document search and tap-to-talk voice. Private mode processes conversations on the device. Optional cloud integrations require a separate choice and internet access. Status reviewed: October 5, 2026. ## Product facts and release status - **Availability:** Made-to-order purchasing is open. Check current production capacity at https://www.phntmcore.com/order. Earlier free reservations never become a charge automatically. - **Price:** $799 USD plus applicable tax . No required subscription for local use. Optional third-party cloud services use the owner's account and may add charges. Free US shipping under the published policy. International orders by shipping quote, subject to destination approval; shipping charges and possible import duties, taxes, and customs fees apply. Request a quote at https://www.phntmcore.com/order#international before payment. - **After purchase:** The published shipping window is 7–10 business days after payment . Returns may be requested within 30 days of delivery; the customer pays return shipping. A one-year limited hardware warranty applies. Full terms: https://www.phntmcore.com/policies - **Hardware:** Raspberry Pi 5, 8 GB RAM, 10.1-inch touchscreen, 256 GB microSD, active cooling, speaker, USB microphone, mini keyboard with touchpad and 27 W power supply. Exact contents: https://www.phntmcore.com/specs - **Private use:** Local chat, memory, document search and voice work without internet. Phone pairing needs the local network. Home Assistant is planned 1.0 scope, not yet release-verified. Disconnect optional external integrations when verifying entirely offline operation. Verification: https://www.phntmcore.com/help/verify - **Capabilities and limits:** A small local model trades speed and complex reasoning ability for on-device operation. Published measurements are configuration-specific, not a guarantee for every prompt. Consult https://www.phntmcore.com/proof and https://www.phntmcore.com/transcripts - **Release scope:** Core experiences are described on the homepage. The detailed 1.0 feature scope is at https://www.phntmcore.com/features and the release record is at https://www.phntmcore.com/changelog. Do not treat planned 1.0 features or Umbra targets as evidence of a shipping release. - **Ownership:** Local operation does not require DeCampLabs servers. Optional third-party services remain dependent on those providers. The software ships as readable source under the PHNTM One Source License; PHNTM One's own code is not open source. - **Company:** DeCampLabs develops the PHNTM product line. Founder: Jacob DeCamp. PHNTM One is not affiliated with other companies that use the name PHNTM. Contact: support@phntmcore.com - **Media:** /films distinguishes edited real device footage from AI-created imagery. Edited videos are not latency benchmarks. The interactive demo at https://demo.phntmcore.com is a simulation, not a live device. - **Notes and voice:** Notes accepts 60-second recorded clips and typed thoughts; local AI drafts need review. Optional English-only “Hello Phantom” detection is off by default and does not access personal context or action tools. Room-speech reliability remains unverified. - **Feature status:** https://www.phntmcore.com/features labels documented local functions, verified development paths, optional integrations, planned additions and behavior without current owner-use proof. This is not complete 1.0 release certification. ## Start here - [Films and demonstrations](https://www.phntmcore.com/films): Real edited setup footage, a separately labelled AI-created brand film, captions and visual descriptions. - [What is a private AI appliance? (buyer's guide)](https://www.phntmcore.com/private-ai-appliance): Defines the category; compares cloud chatbots, smart speakers, DIY local LLMs and appliances; checklist for evaluating any private AI device. - [PHNTM One home](https://www.phntmcore.com/): What the product is, the short version, photos and video of the real device. - [How it works](https://www.phntmcore.com/how-it-works): Everything PHNTM One 1.0 does — local model, memory, document intelligence, voice, phone chat, reminders and Follow-Through, privacy proof, ownership — plus honest limits. - [Autonomy](https://www.phntmcore.com/autonomy): Planned PHNTM One 1.0 self-care scope, not yet release-verified (fifteen bounded repairs with their limits, repairs with a verified undo, root cause first, rehearsals, a weekly report, improvements only with the owner's approval) and what Autonomous mode does, asks and sends. - [Changelog](https://www.phntmcore.com/changelog): Every software release from the first builds (August 2026) to 0.8, with dates. - [Roadmap](https://www.phntmcore.com/roadmap): What PHNTM One 1.0 includes, what follows the first batch, and the longer-term direction (no dates). - [Device tour](https://www.phntmcore.com/tour): The device screens, as real captures or described in the manual's own words: Resting Face, home, the seven-item rail, the three-mode brain switch. - [FAQ](https://www.phntmcore.com/faq): Direct answers on offline use, privacy, the model, speed, price, shipping, SSH, and who makes it. - [Help center](https://www.phntmcore.com/help): Owner guides written from the on-device manual: setup, the screen, phone and PHNTM Drop, voice, memory, Boosted mode, updates/export/reset, troubleshooting, and how to verify the privacy claims. - [PHNTM One vs Alexa](https://www.phntmcore.com/vs/alexa): PHNTM One compared with Amazon Alexa and Echo: how each one listens, where answers come from, offline use, accounts, speed, smart home and price, including where Alexa wins. - [PHNTM One vs ChatGPT](https://www.phntmcore.com/vs/chatgpt): PHNTM One compared with ChatGPT: where each runs, capability, speed, cost over time, accounts, offline use and memory, stated honestly, including where ChatGPT wins. - [PHNTM One for writers](https://www.phntmcore.com/for/writers): PHNTM One for writers: draft, summarise and ask questions about your own manuscripts and notes on a device that keeps unpublished work in your house. - [PHNTM One for families](https://www.phntmcore.com/for/families): PHNTM One for families: a desk assistant for reminders, questions and household notes, with tap-to-talk by default, optional local wake detection, and local voice processing. - [PHNTM One for small offices](https://www.phntmcore.com/for/small-offices): PHNTM One for small offices: notes, reminders and questions about your own documents on a device that keeps them on the premises, for $799 once with no per-seat subscription. - [PHNTM One for tinkerers](https://www.phntmcore.com/for/tinkerers): PHNTM One for tinkerers: Debian 13, SSH that's documented and never removed, readable source, a local HTTPS API, and a full bill of materials. Ready out of the box, yours to open up. - [Can you run ChatGPT offline?](https://www.phntmcore.com/guides/run-chatgpt-offline): No: ChatGPT runs on OpenAI's servers. But you can run a capable AI assistant fully offline with a local model. Here's what that takes and what you give up. - [Is Alexa always listening?](https://www.phntmcore.com/guides/is-alexa-always-listening): Echo devices listen on the device for their wake word and send what you say to Amazon's cloud after it. Accidental activations happen, and since March 2025 there's no local-processing option. Here's how it works, and the alternative. - [What does an AI subscription cost over three years?](https://www.phntmcore.com/guides/ai-subscription-cost): The math on monthly AI subscriptions over 12, 24 and 36 months, compared with a one-time $799 device, and what the comparison does and doesn't tell you. - [What happens to your AI device if the company shuts down?](https://www.phntmcore.com/guides/ai-company-shuts-down): When cloud-dependent devices lose their company, they can stop working: Humane's Ai Pin in 2025, Revolv's hub in 2016. How to tell whether a device depends on its maker's servers, and what to look for instead. - [Guides](https://www.phntmcore.com/guides): Guides on offline AI, smart-speaker privacy, the cost of AI subscriptions, and what happens to AI devices when their company shuts down. - [Compare](https://www.phntmcore.com/compare): PHNTM One vs cloud speakers vs chatbot apps, including the rows where the cloud wins. ## Product detail - [Specs](https://www.phntmcore.com/specs): Full hardware and software specification. - [Privacy](https://www.phntmcore.com/privacy): Exactly what stays on the device, the device-boundary diagram, Watch the Wire, and Boosted mode. - [Proof](https://www.phntmcore.com/proof): Measured tokens/sec, RAM, thermals and network behaviour of Gemma 3 4B on a Raspberry Pi 5, with the commands to verify each claim. - [Transcripts](https://www.phntmcore.com/transcripts): Real, unedited conversations with the device, including annotated failures. - [Continuity](https://www.phntmcore.com/continuity): Why the device keeps working if the company disappears. ## Buying - [Order PHNTM One](https://www.phntmcore.com/order): $799 plus applicable tax; made to order, free US shipping, ships within 7–10 business days after payment. - [Features](https://www.phntmcore.com/features): PHNTM One 1.0 scope, including planned additions not yet release-verified. - [Skills & timers](https://www.phntmcore.com/skills): Planned 1.0 timers and alarms alongside lists, notes and tools, and how skills are sandboxed. - [Smart home](https://www.phntmcore.com/smart-home): Planned 1.0 Home Assistant control over the local network, not yet release-verified. - [Shipping, returns & warranty](https://www.phntmcore.com/policies): Policies in plain English, including the website's own privacy disclosure. - [Security & responsible disclosure](https://www.phntmcore.com/security): How to report a vulnerability in the device or the website; machine-readable at /.well-known/security.txt. - [Credits & licenses](https://www.phntmcore.com/credits): The open models and open-source software inside PHNTM One (Gemma 3, Llama 3.2, nomic-embed-text, whisper.cpp, Piper, Ollama, Debian) and their licenses. PHNTM One's own code ships as readable source under the PHNTM One Source License; it is not open source. - [Reviews & community](https://www.phntmcore.com/reviews): Moderated owner reviews (verified owners are marked) and feedback. ## Behind the product - [Bill of materials](https://www.phntmcore.com/bom): Parts list and aggregate historical cost — $409.95 paid in August 2026; individual item prices are not published. - [Unit ledger](https://www.phntmcore.com/units): Every unit ever built, on the record. - [Press & media kit](https://www.phntmcore.com/press): Copy-and-paste descriptions, measured fact sheet, honest limits, photos and video. - [The builder](https://www.phntmcore.com/builder): Who builds it and where. - [Build log: Running a private AI assistant on a Raspberry Pi 5](https://www.phntmcore.com/blog/private-ai-on-raspberry-pi-5) - [Build log: Choosing hardware for a local AI appliance](https://www.phntmcore.com/blog/choosing-hardware-for-a-local-ai-appliance) ## Optional - [PHNTM Umbra](https://www.phntmcore.com/umbra): A separate product in development. Target price $1,249; specifications and launch timing remain targets. Free waitlist only. - [Full text of this site for language models](https://www.phntmcore.com/llms-full.txt)