What a $6,880 AI Agent Actually Does (And What It Can't)
By Warren Schuitema, Founder | Matchless Marketing | The AI Dad
Vertu just put a price tag on an AI agent, and it's $6,880. That number is worth paying attention to, not because you should buy the thing, but because of what it reveals about where AI agents actually stand right now.
The product is the Vertu Alphafold, a luxury foldable phone running a built-in AI agent called Hermes Agent. Hermes is a pre-installed AI agent built on top of the open-source Hermes project. Vertu says it can analyze files, automate tasks across apps, remember conversations, and hand off requests to a human concierge when needed. Unlike most smartphone AI assistants that largely just respond to prompts, Hermes is designed to execute multi-step workflows on users' behalf. On paper, it's the closest thing yet to an AI agent marketed as a serious business tool, not a feature buried in your settings.
TechCrunch put it through real-world testing this week. The results are instructive for anyone thinking about AI agents, at any price point.
What the Vertu Alphafold Actually Is
The Alphafold starts at $6,880, and the review unit came wrapped in genuine calfskin leather with titanium accents, setting it apart from mainstream foldables that largely rely on glass or synthetic finishes. Higher-end variants go further. Those models feature bespoke finishes including alligator leather, 18K gold, and natural diamond accents. The top-end standard model is priced at $46,800.
Here's what they don't put in the press release: TechCrunch noticed striking similarities between the device and the $1,100 ZTE Nubia Fold, from the hinge design and dimensions to the placement of the speakers, microphones, and the fingerprint reader. Vertu confirmed the Alphafold was developed through a specialist supply-chain partnership involving ZTE/Nubia's hardware platform. The company said it was responsible for the luxury materials, software experience, quality control, and after-sales service.
So the hardware you're mostly paying for is available at a fraction of the price. The real bet Vertu is making is on Hermes Agent.
What Hermes Agent Can Actually Do
The Alphafold can route requests across multiple AI models, including OpenAI's GPT, Anthropic's Claude, Google's Gemini, and selected open-source models, while also integrating with more than 80 apps and dozens of native phone functions for cross-platform workflows. The agent can connect to enterprise systems like ERP and CRM, and coordinate tasks such as approvals, scheduling, sales tracking, travel planning, and operational reporting through natural-language prompts.
In testing, some of that held up. Hermes impressed when analyzing local files and spreadsheets, areas where Gemini on Samsung's foldable still relied on manually uploaded documents. It was also more willing to automate actions across apps and complete multi-step workflows.
That's real. File analysis without manual uploads is a genuine workflow win, especially for anyone reviewing contracts, quotes, or financial summaries on the go.
But the autonomy has limits. One test replicated a typical executive situation just before heading to the airport. The assistant dispatched a text, activated Do Not Disturb mode, and launched Google Maps with the airport set as the target. However, it failed to start navigation on its own, and set a reminder for the wrong time entirely.
That's not a minor bug. That's an agent making confident decisions that were just wrong. No clarification asked, no confirmation sought, just incorrect action taken.
Early software builds also struggled to upload files, analyze images, and connect to Vertu's concierge service. After these issues were reported, the company rolled out server-side fixes mid-review. For a $6,880 device, patching functionality during the review process is not a great look.
The Security Question Nobody Can Answer Yet
Vertu pitches the Alphafold as a privacy-first device. The device features a proprietary A5 security chip designed to isolate authentication keys, biometric credentials, and sensitive enterprise information from the main operating system. Commercially sensitive data can be processed locally, while prompts sent to external AI models are redacted or tokenized before leaving the phone.
That architecture sounds solid. The problem is that it hasn't been verified by anyone independent. The system has not yet undergone third-party security audits or independent certification. Vertu told TechCrunch that independent audits remain on its security roadmap as "an explicit next-stage commitment."
For any small business owner or executive running client data through an AI agent, that matters. Security claims without audits are just claims.
What This Tells You About AI Agents Right Now
Here's my honest read on this, and it's the reason I wanted to write about it.
The Vertu story is not really about Vertu. Most of us aren't buying a $6,880 phone. But the performance gaps Hermes Agent showed, acting autonomously but incorrectly, needing mid-review patches, struggling with timing and context, those are the exact same gaps you'll hit with any AI agent you set up today, whether it's in n8n, ChatGPT, Claude, or a Zapier workflow.
The Alphafold is an ambitious attempt to build an AI-first luxury smartphone, but the execution falls short of its price tag. Despite its premium materials and exclusive services, the core hardware offers little that cannot be found in significantly less expensive foldables, while Hermes Agent remains an evolving platform rather than a compelling reason to spend thousands more.
That's the honest verdict on AI agents broadly right now. They're evolving platforms, not finished products. The best ones handle structured, bounded tasks reliably. Local file analysis, document summarization, templated follow-ups, calendar blocking. They fall apart when the task requires genuine judgment about context they don't have.
The small business owners I work with who are getting real value from AI agents aren't asking them to "run the business." They're asking them to do one specific thing extremely well. Review this contract and flag any payment terms outside 30 days. Pull my last five invoices and format them into a summary. Draft a follow-up email based on these notes. Bounded. Specific. Reviewable.
The Vertu pitch is that an agent can run your executive day. The TechCrunch review shows it still sends wrong reminders and needs a human to catch its mistakes. That's exactly where we are with agents in mid-2026, for $6,880 or for free.
What You Should Actually Do with This Information
If you've been hesitant to try AI agents because you're waiting for them to be "ready," this story should help calibrate your expectations without killing your interest. They're ready for bounded tasks. They're not ready for autonomous responsibility.
The practical move: pick one repeatable task in your business that has clear inputs and a clear output. Something you do the same way every time. Draft a client proposal from a set of notes. Summarize a weekly sales report. Send a check-in message three days after a quote goes out.
Set up that one workflow, in ChatGPT with a custom GPT, in Claude Projects, or in n8n if you want real automation. Run it for two weeks with your review before anything goes out. That's how you build trust with an agent, the same way you'd build trust with a new hire.
The people selling $6,880 AI agents are betting you'll pay for someone else to have already figured this out. You don't need to. You need a clear task, the right tool, and a habit of checking the work until you trust it.
Start there.