Voltage Park
US neocloud selling on-demand and reserved bare-metal NVIDIA HGX H100 clusters from $1.99/hr; acquired TensorDock (2025) and merged with Lightning AI (Jan 2026).
A real, actively used decentralised inference provider run by a named (offshore) company, with transparent per-token pricing, attestable TEE runs, TAO/Stripe payment and a live status page - the clearest example of buying compute inside Bittensor, though capacity has historically been oversubscribed and there is no uptime SLA.
site reachable · checked 11 h ago · daily check
by the bestcloudmining editorial desk · curated review · prose updated 2026-10-02 · facts as of 2026-10-02 · how we verify
Chutes runs a Bittensor subnet (subnet 64) offering 14 TEE-attested LLM models with published per-token rates, operated by Chutes Global Corp (Nevis) with Rayon Labs as the developer [2][4]. It is the most directly usable product in the Bittensor ecosystem - a real API with real pricing, not a narrative.
LLM inference per token (14 models, all TEE-attested for confidential inference) [1], plus private GPU deployments (RTX Pro 6000 Blackwell at $1.80/hour plus a one-time $5.40 deployment fee) [1]. Payment: TAO deposits or Stripe [2]. A new Base plan at $3/month exists [2]. There is no per-GPU-hour rental in the traditional sense - pricing is per 1 million input/output tokens for inference [2].
The contracting entity is Chutes Global Corp (Nevis) [3]. Terms contain no KYC section [3]. Crypto payments are final and non-refundable [4]. Stripe payments may carry standard card protections [3]. No SLA is published [3]. The privacy policy covers API usage and payment data [3].
The models API is public - 14 models with per-model pricing, all marked confidential_compute: true (TEE-attested) [4]. SN64 trades at about $21.82 with a market cap of over 134 million dollars [6] [5]. The status page showed 100% uptime at check [2]. What stays unverifiable: actual request volumes and developer counts (only "trillions of tokens/month" self-reported), and the TEE attestation's independent audit status [4].
The product is live with 100% uptime on the status page [2]. A Kraken listing (July 2026) is documented [4]. The Chutes Global Corp / Rayon Labs split is explicit in the terms [3]. No payout complaints or fraud patterns surfaced [3]. The "trillions of tokens/month" claim is self-reported, not independently verified [2].
Published per-token rates: Mistral-Nemo $0.0245/$0.0978 per 1M tokens up to Kimi K3 at $3.00/$15.00 [1]. Private RTX Pro 6000: $1.80 per GPU-hour + $5.40 one-time deployment fee [1]. Base plan: $3/month [2]. SN64 token: about $21.82, market cap of over 134 million dollars [6]. Payment: Stripe (card) or TAO deposits [3]. Crypto final, non-refundable [4]. KYC: not required [3].
Chutes is the most usable consumer-facing Bittensor product - real per-token pricing, TEE-attested models, and a straightforward API [1][2]. The Nevis-incorporated entity and non-refundable crypto policy are the honest caveats [3]. For LLM inference with privacy, it is a verifiable option [1]; for GPU rental, the private-deployment path costs more than commodity clouds but adds TEE attestation [1]. The Bittensor integration means the subnet earns TAO emissions that subsidize the inference cost [2] - the $3/month Base plan is priced for accessibility rather than profit [2]. If you are comparing Chutes against centralized inference APIs (OpenAI, Anthropic), the TEE-attestation is the differentiator that centralized clouds do not offer [4]. The 14-model catalog rotates as models are added and removed - check the live API for current availability [4]. The crypto-only refund policy is the honest trade-off for the decentralized model [3]. Chutes is one of the few Bittensor subnets with a clear consumer product, published rates, and a named legal entity - which is why it scores highest among the subnet-based services [2].
| Documented price | $0.025 per GPU/hr · $1.80 per GPU/hr (private TEE RTX Pro 6000; per-1M-token input starts at $0.0245) |
|---|---|
| Notes | Per-1M-token public inference (observed 2026-09-24) starts at $0.0245 in / $0.0978 out (Mistral-Nemo, Nemotron-3-Nano-Omni-30B), up to Kimi K3 at $3.00/$15.00; 14 TEE models via GET https://llm.chutes.ai/v1/models. Private self-serve chutes are billed per second at the GPU hourly rate: the only listed TEE class, RTX Pro 6000 96GB, is $1.80/hr plus a one-time deploy fee of 3x the hourly rate ($5.40). |
Run the long math before paying: cloud mining calculator →
Built only from what is documented on this page as of 2026-10-02 - anything not documented stays a gap on purpose.
Everything happens on chutes.ai - we count the click, nothing else changes for you.
Price from $0.025 per GPU/hr - start with the minimum, not with the plan the checkout page pushes.
Accepted: TAO plus card transfer. Check the network and the deposit address inside your account, never from an email.
maintenance · refund: Cryptocurrency payments are final and non-refundable; Stripe payments may be refunded only in limited…. A term you did not read is money you cannot get back.
Proof is documented as yes - Models run in Trusted Execution Environments with confidential_compute and attestation. Compare the first payout with the pool or wallet before scaling up.
Then run the math: mining calculator · contract vs. HODL · all GPU compute providers.
Short answer: Based on our checks as of 2026-10-02: Chutes (Bittensor subnet) scores 6.7/10 in our weighted review. We logged no red flags. User reports and operator background are documented with dated sources in the meta-review below.
| Status | active |
|---|---|
| Operator | Chutes · HQ Saint Kitts and Nevis (high) |
| Founded | 2024 |
| Proof | published |
| KYC | not documented |
No red flags logged as of 2026-10-02.
A real, operating provider can still lose you money: with a contract you own no hardware, revenue follows coin price and difficulty, and most published contracts are priced so the seller comes out ahead. "Legit" means the operator exists and pays what its terms promise - not that the deal is profitable. 12 dated sources back this page; they are listed under sources.
Same category, scored on the same criteria - so the comparison stays honest. Compare only in the same unit, and read the minimum, KYC and proof columns before switching.
US neocloud selling on-demand and reserved bare-metal NVIDIA HGX H100 clusters from $1.99/hr; acquired TensorDock (2025) and merged with Lightning AI (Jan 2026).
Distributed consumer-GPU cloud with published per-second prices and measured host earn rates; no token, fiat only.
Established US GPU cloud: on-demand H100/B200 instances, 1-Click Clusters and Superclusters, billed per minute with no ingress/egress charges.
Open permissionless marketplace for containerised GPU/CPU compute; providers bid, tenants pay per block.
Ready to try Chutes (Bittensor subnet)? Start small, verify one payout, and only scale up once the hashrate or compute actually matches what you paid for.
Based on our checks as of 2026-10-02: Chutes (Bittensor subnet) scores 6.7/10 in our weighted review. We logged no red flags. User reports and operator background are documented with dated sources in the meta-review below.
No comments yet - be the first. Posting runs a short proof-of-work in your browser (anti-spam), no account needed.