When researching decentralized and aggregated cloud providers, many independent hardware owners stumble across TensorDock. With aggressive renter pricing, an extensive catalog spanning from consumer RTX cards to enterprise H100 SXM5 nodes, and flexible pay-per-second virtual machine billing, it often appears on shortlists for compute providers.
However, if you are an individual operator looking to monetize a desktop rig or a small multi-GPU setup in your home or garage, you will quickly hit a wall.
Unlike open peer-to-peer marketplaces, TensorDock operates under strict commercial vetting and infrastructure standards. This review examines TensorDock’s host model, its 25% revenue-share structure, and why it is built exclusively for datacenters rather than residential providers. For a comparison of the open marketplaces (Vast.ai and Clore.ai), read our full comparison guide.
💡 Quick Answer: No, you cannot easily list a home PC on TensorDock. The platform strictly requires hardware to be housed in verified commercial datacenters or Tier-3/Tier-4 colocation facilities, backed by enterprise-grade networking and strict 99.99% uptime SLAs.
1. What Is TensorDock? (The Marketplace Model)
Founded as a cloud broker and compute marketplace, TensorDock aggregates virtualized and bare-metal server capacity from independent infrastructure providers around the globe.
For tenants and renters, it offers an efficient API-driven cloud where they can spin up KVM virtual machines equipped with GPUs like the RTX 4090, A100, or H100 at rates significantly lower than traditional hyperscalers (AWS, Azure, GCP). The appeal is instant provisioning, per-second billing, and a wide selection of hardware tiers. Unlike hyperscalers that lock you into long-term contracts, TensorDock allows developers to rent a GPU for 10 minutes, run a specific inference task, and shut it down instantly.
For suppliers, TensorDock handles client billing, user onboarding, frontend provisioning, and payment distribution, leaving operators free to manage their physical server racks. In essence, they are the “middleware” between enterprise-grade hardware owners and AI developers who need compute. They abstract away the complexity of the marketplace, but they demand professional-grade reliability in return.
2. TensorDock Host Requirements: Why Home Rigs Are Rejected
While open networks like Vast.ai welcome individual hobbyists with residential internet connections, TensorDock enforces rigorous supplier compliance:
- Commercial Facility Mandate: New hardware must be deployed inside an established data center or a professionally managed colocation facility. Residential IP blocks and domestic broadband jitter do not pass their vetting procedures. If your server is in a bedroom, it won’t be accepted. The platform checks for Tier-3 or Tier-4 colocation standards, which guarantee redundant power (N+1), redundant cooling, and physical security.
- Strict Uptime SLAs: TensorDock holds its infrastructure hosts to a 99.99% uptime standard. Nodes failing to meet reliability thresholds are systematically removed from the marketplace. This requires redundant power (UPS), redundant internet (multi-homed BGP), and 24/7 on-site support. In a residential setting, a simple power outage or ISP drop would violate this SLA instantly.
- Enterprise Networking: Hosts must provide dedicated IP addresses or fully forwarded port blocks capable of sustaining rigorous multi-gigabit throughput under heavy renter loads. A standard home router cannot handle this. You need enterprise-grade switches, BGP routing, and proper VLAN segmentation to isolate tenant traffic.
👉 Review which platforms actually welcome individual operators on our Host Eligibility Matrix.
3. The Financial Structure: 25% Revenue Share
For those operating qualifying commercial datacenters, TensorDock’s economic model relies on a flat revenue-split arrangement:
- Revenue Share Fee: TensorDock takes a 25% platform cut from gross earnings, leaving the infrastructure host with 75%. This is comparable to the markup Vast.ai applies to renter prices, but here it is explicitly stated as a revenue share.
- Payout Handling: Payouts are distributed via Stripe once minimum balance thresholds are met. This means you get fiat USD, not volatile crypto tokens, which is a major advantage for professional operators.
- No Referral Program: Unlike some P2P networks that offer host referral bonuses, TensorDock does not currently operate a public host onboarding referral program, per their corporate disclosures. They rely on organic supply growth and word-of-mouth from datacenter operators.
For a datacenter operator, this 25% cut is comparable to what Vast.ai effectively takes via its markup. The difference is that TensorDock provides a managed, enterprise-grade frontend, whereas Vast.ai provides a raw marketplace. TensorDock handles the billing, the UI, the customer support, and the legal compliance. You just provide the metal.
👉 Compare these take-rates against consumer-focused networks in our Electricity Cost Ceiling Guide.
4. TensorDock vs. Open P2P Networks (Vast.ai & Clore.ai)
To understand where TensorDock fits in the ecosystem, it helps to compare it directly against open alternatives:
| Evaluation Metric | TensorDock | Vast.ai / Clore.ai |
|---|---|---|
| Target Host | Datacenter & Colocation Operators | Home Hosts, Miners, & Individuals |
| Hardware Vetting | Strict (Tier-3/4 facility checks) | Relaxed (Daemon verification) |
| Platform Cut | 25% Revenue Share | ~25% Markup (Vast) or 5% Base (Clore) |
| Uptime Expectation | 99.99% Enterprise SLA | Flexible / Algorithmic Ranking |
| Our Site Role | Renter Price Benchmark Only | Active Host Calculator Integration |
Because TensorDock is locked to commercial infrastructure, we display its pricing strictly as a Market Renter Benchmark rather than an option in our host earnings calculator. It is a reference point for what enterprise buyers pay, but it is not a viable path for the average home host. For a look at how enterprise-grade servers compare in cost, read our Enterprise AI Server Guide.
If you are a home host, you should focus on Vast.ai and Clore.ai. If you own a datacenter, TensorDock is a viable channel to monetize your unused capacity without building your own marketplace.
👉 Check current comparative market rates on our Market Rates Table.
5. Frequently Asked Questions
Can I host a GPU on TensorDock from my house?
No. TensorDock requires hosts to operate out of verified commercial datacenters or colocation facilities with enterprise networking and strict uptime guarantees. Residential setups are rejected during onboarding.
Does TensorDock offer a referral program for hosts?
No official host referral or onboarding commission program is offered by TensorDock, unlike several competing peer-to-peer platforms.
How does TensorDock compare to Vast.ai for earnings?
TensorDock takes a flat 25% revenue share and caters to professional operators, whereas Vast.ai operates an open P2P marketplace where individual home rigs can freely compete. If you have a home rig, Vast.ai is the only viable option.
Where can I check which platforms allow home hosting?
Review our verified onboarding rules and eligibility breakdowns directly on the Host Eligibility Matrix.
What is the uptime requirement for TensorDock hosts?
TensorDock enforces a strict 99.99% uptime SLA. Nodes that fail to meet this reliability threshold are systematically removed from the marketplace.
6. Sources & Official Documentation
- TensorDock Host Operator Portal — infrastructure and facility mandates (checked October 2026)
- TensorDock Corporate Information — platform terms and program details (checked October 2026)
- Comparative marketplace telemetry: GPUEARN Market Rates Table
- Individual host onboarding guidelines: GPUEARN Host Eligibility Matrix
Written by Alex Rivers (Senior Hardware Infrastructure Analyst) and the GPUEARN Research Team. Institutional compute requires institutional infrastructure.
Last updated: October 25, 2026