{ "features_resolved": { "snapshot_any": false, "snapshot_mem": false, "fork": false, "pause_resume": true, "persistent_disk": true, "whole_disk": null, "persist_scope": null, "volumes": true, "auto_stop_idle": false, "long_sessions": true, "custom_image": null, "image_snapshot": false, "image_oci": false, "vm_isolation": true, "docker_inside": true, "nested_virt": false, "root": true, "systemd": null, "browser": false, "desktop": false, "computer_use": false, "code_interpreter": false, "browser_control": null, "desktop_control": null, "browser_and_desktop": false, "anti_bot": false, "captcha_solving": false, "residential_ip": false, "agent_harnesses": false, "ssh": true, "public_ipv4": true, "inbound_https": false, "custom_domain": false, "raw_tcp_inbound": false, "egress_allowlist": false, "open_internet": true, "static_egress_ip": false, "private_network": null, "self_host": false, "byoc": false, "open_source": false, "soc2": true, "hipaa": true, "sso": false, "gpu": true, "arm64": false, "windows": false, "macos": false, "lang_python": true, "lang_node": false, "lang_go": null, "lang_rust": null, "lang_java": null, "gpu_desktop": false, "credential_injection": null, "egress_http_rules": false, "wake_on_request": false, "live_resize": false, "fork_running": false, "memory_snapshot_fork": false, "webhooks": null, "scheduled_wakeups": null, "audit_logs": null, "eu_data_residency": false, "zero_data_retention": null, "scoped_api_keys": null, "spend_limits": null, "mcp_server": false, "usage_api": true, "snapshot_auto": false, "snapshot_on_demand": false, "harness_api": false, "own_agent_api": false, "fast_boot": null, "strong_isolation": true, "ingress_rules": true, "guest_firewall": true, "secret_proxy": false, "secret_proxy_any": false, "volume_attach": true, "volume_shared": true, "sidecar": null, "fuse": true, "kernel_features": true, "burst": null }, "company": { "id": "voltage-park", "checked": "2026-09-30", "company": "Voltage Park", "ownership": { "value": "subsidiary", "parent": "Navigation Fund (nonprofit founded by Jed McCaleb, co-founder of Stellar/Ripple/Mt. Gox and Vast)", "source": "https://techcrunch.com/2023/10/31/new-nonprofit-backed-by-crypto-billionaire-scores-ai-chips-worth-500m/", "quote": "The Navigation Fund then purchased the aforementioned H100s ... and transferred ownership of the GPUs to for-profit Voltage Park, which is technically a subsidiary of the Navigation Fund." }, "stage": { "value": "subsidiary", "source": "https://techcrunch.com/2023/10/31/new-nonprofit-backed-by-crypto-billionaire-scores-ai-chips-worth-500m/", "quote": "Voltage Park, which is technically a subsidiary of the Navigation Fund.", "confidence": "high", "note": "Not VC-funded in the traditional sense: capitalized by a one-time ~$500M GPU asset contribution from its nonprofit parent, the Navigation Fund (funded by Jed McCaleb), rather than priced equity rounds." }, "accelerators": [], "total_funding_usd": { "value": 500000000, "status": "confirmed", "basis": "TechCrunch: Navigation Fund purchased 24,000 Nvidia H100 GPUs (~$500M, 'worth an estimated half a billion dollars') and contributed them as initial capital to Voltage Park. Voltage Park's own About page separately claims a '$1 billion Long-term capital foundation', which is unconfirmed by press and may include later debt/hardware financing not itemized publicly.", "source": "https://techcrunch.com/2023/10/31/new-nonprofit-backed-by-crypto-billionaire-scores-ai-chips-worth-500m/", "quote": "Already, the cluster of GPUs — worth an estimated half a billion dollars and among the largest in the world — is being used by startups Imbue and Character.ai" }, "last_round": { "type": "capital contribution (nonprofit-to-subsidiary GPU asset transfer, not a priced equity round)", "amount_usd": 500000000, "date": "2023-10", "leads": [ "Navigation Fund (Jed McCaleb)" ], "status": "confirmed", "source": "https://techcrunch.com/2023/10/31/new-nonprofit-backed-by-crypto-billionaire-scores-ai-chips-worth-500m/", "quote": "McCaleb started a nonprofit called the Navigation Fund to whose endowment he donated. The Navigation Fund then purchased the aforementioned H100s ... and transferred ownership of the GPUs to for-profit Voltage Park" }, "investors": [ "Navigation Fund", "Jed McCaleb" ], "portfolio_listings": [], "revenue": { "value": null, "status": "unknown" }, "employees": { "value": 80, "source": "https://www.voltagepark.com/about" }, "note": "Voltage Park is a for-profit subsidiary of the nonprofit Navigation Fund (funded by crypto/aerospace entrepreneur Jed McCaleb), launched Oct 2023 to operate a ~24,000-GPU (Nvidia H100) cloud; McCaleb does not own, run, or profit from either entity per a Navigation Fund spokesperson. Former CEO Ozan Kaya has since moved to become President of Lightning AI, per Voltage Park's team page, suggesting a leadership/organizational change worth confirming with a buyer. No subsequent priced equity round or updated total-funding figure beyond the initial ~$500M GPU contribution was found; the company's own About page claims a '$1 billion long-term capital foundation' which is unverified by independent press." }, "id": "voltage-park", "name": "Voltage Park", "url": "https://www.voltagepark.com/pricing", "category": "gpu-cloud", "isolation": "vm", "modes": [ { "key": "on-demand-h100", "label": "HGX H100 on-demand (Ethernet), from $1.99/GPU-h", "flags": [], "os": [ "linux" ], "regions": [ "us" ], "gpu_only": true, "gpu_min_count": null, "features": {}, "pricing": "resource", "vcpu_h": 0, "ram_gib_h": 0, "cpu_basis": "alloc", "ram_basis": "alloc", "active_floor": 0, "ram_per_vcpu": [ null, null ], "ram_fixed_per_vcpu": null, "vcpu_options": null, "min_vcpu": null, "max_vcpu": null, "max_ram_gib": null, "sizes": null, "pool_tiers": null, "requires_always_on": false, "always_on_month_per_instance": null, "commit_note": null, "min_billed_seconds": 3600, "granularity_s": null, "start_fee": 0, "boot_overhead_s": null, "multiplier": 1, "windows_vcpu_h": null, "gpu": { "H100": 1.99 }, "note": "Added 2026-09-28 (missing-providers audit). https://www.voltagepark.com/pricing: FAQ \"You can rent one H100 GPU for 1 hour starting at $1.99 without a contract\"; on-demand Ethernet 1-1016 HGX H100 GPUs, self-serve, no minimum term. Tiles say \"Contact for pricing\"; $1.99 is a \"starting at\" figure. Bundled vCPU/RAM per GPU not published (max_vcpu/max_ram null). Billing increment not published (hourly assumed)." }, { "key": "on-demand-h100-ib-8x", "label": "HGX H100 on-demand InfiniBand (8+ GPUs)", "flags": [ "sales" ], "os": [ "linux" ], "regions": [ "us" ], "gpu_only": true, "gpu_min_count": 8, "features": {}, "pricing": "resource", "vcpu_h": 0, "ram_gib_h": 0, "cpu_basis": "alloc", "ram_basis": "alloc", "active_floor": 0, "ram_per_vcpu": [ null, null ], "ram_fixed_per_vcpu": null, "vcpu_options": null, "min_vcpu": null, "max_vcpu": null, "max_ram_gib": null, "sizes": null, "pool_tiers": null, "requires_always_on": false, "always_on_month_per_instance": null, "commit_note": null, "min_billed_seconds": null, "granularity_s": null, "start_fee": 0, "boot_overhead_s": null, "multiplier": 1, "windows_vcpu_h": null, "gpu": null, "note": "Added 2026-09-28 (missing-providers audit). https://www.voltagepark.com/pricing: on-demand 3200 Gbps InfiniBand, 8-1016 HGX H100, \"Contact for pricing\"." } ], "storage": { "included_disk_gib": 0, "disk_gib_month": null, "disk_billed_when_stopped": false, "snapshot_gib_month": null, "paused_ram_billed": false }, "network": { "egress_free_gib": null, "egress_gib": 0, "ipv4_month": null, "ipv4_available": null, "internet": { "default": "full" } }, "plans": [ { "name": "On-demand", "fee": 0, "flags": [], "fee_is_credit": false, "included_usd": 0, "per_seat": 0, "waives_start_fee": false, "included_applies_to_modes": null, "concurrency": null, "max_session_h": null, "max_vcpu": null, "max_ram_gib": null, "sizes_allowed": null, "note": "Self-serve, no minimum term." }, { "name": "Long-term reserve", "fee": null, "flags": [ "sales" ], "fee_is_credit": false, "included_usd": 0, "per_seat": 0, "waives_start_fee": false, "included_applies_to_modes": null, "concurrency": null, "max_session_h": null, "max_vcpu": null, "max_ram_gib": null, "sizes_allowed": null, "note": "6+ month contracts, 32-8000+ H100s, contact sales." } ], "free": { "monthly_credit": 0, "one_time_credit": 0 }, "price_expires": null, "features": { "linux": true, "windows": false, "macos": false, "arm64": false, "gpu": true, "gpu_types": [ "H100" ], "snapshot": "none", "pause_resume": true, "fork": false, "persistent_disk": true, "volumes": true, "auto_stop_idle": false, "max_session_h": null, "custom_image": null, "docker_inside": true, "nested_virt": false, "root": true, "systemd": null, "browser": false, "desktop": false, "computer_use": false, "code_interpreter": false, "agent_harnesses": false, "ssh": true, "public_ipv4": true, "inbound_https": false, "custom_domain": false, "raw_tcp_inbound": false, "egress_allowlist": false, "static_egress_ip": false, "private_network": null, "self_host": false, "byoc": false, "open_source": false, "regions": [ "us" ], "soc2": true, "hipaa": true, "sso": false, "lang_python": true, "lang_node": false, "lang_go": null, "lang_rust": null, "lang_java": null, "gpu_desktop": false, "credential_injection": null, "egress_http_rules": false, "wake_on_request": false, "live_resize": false, "fork_running": false, "memory_snapshot_fork": false, "webhooks": null, "scheduled_wakeups": null, "audit_logs": null, "eu_data_residency": false, "zero_data_retention": null, "scoped_api_keys": null, "spend_limits": null, "mcp_server": false, "snapshot_auto": false, "snapshot_on_demand": false, "harness_api": false, "ingress_rules": true, "guest_firewall": true, "secret_proxy": false, "volume_attach": true, "volume_shared": true, "isolation": "qemu-kvm", "image_oci": false, "anti_bot": false, "captcha_solving": false, "egress_ip_type": "datacenter", "own_agent_api": false, "max_session_unlimited": true, "open_internet": true, "usage_api": true }, "perf": { "cpu_runs_s": null, "cold_start_ms": null, "burst_tti_ms": null, "dax_s": null }, "perf_source": "No benchmark row: provider is not in ComputeSDK results or the HPC run (as of 2026-09-28). See research/perf-map.json.", "knobs": [], "caveats": [ "Only a \"starting at $1.99\" per-GPU-hour figure is published (FAQ); tiles say contact for pricing.", "Per-GPU vCPU/RAM and billing increment not published.", "Voltage Park has merged with Lightning AI (site banner, 2026-09-28).", "\"No hidden ingress, egress or support costs\" per pricing page." ], "sources": [ "https://www.voltagepark.com/pricing" ], "audit_2026_09_29": { "applied": 3, "source": "checked against the provider's live pricing and docs pages" }, "feature_evidence": { "snapshot_auto": { "value": false, "source": "https://support.voltagepark.com/article/my-instances", "quote": "Shut Down: Gracefully power off the instance while retaining its data.", "note": "Stop keeps the disk (persistent_disk, and stopped instances bill at the full running rate). No snapshot concept anywhere in the docs or the full On-Demand OpenAPI spec. VMs can additionally 'Deallocate compute resources' (storage-only billing), still disk only.", "checked": "2026-09-29" }, "snapshot_on_demand": { "value": false, "source": "https://cloud-api.voltagepark.com/api/v1/openapi.json", "quote": null, "note": "The On-Demand API (VMs, bare metal, storage, billing, org, locations) has no snapshot/image endpoints; the help center never mentions snapshots. Only a Managed Kubernetes add-on restore object with a 'snapshot_captured' timestamp (cluster-level, not machine snapshots).", "checked": "2026-09-29" }, "harness_api": { "value": false, "source": "https://support.voltagepark.com/article/api", "quote": "Our Voltage Park On Demand API allows developers to deploy, manage, and monitor virtual machines (VMs) and bare metal instances.", "note": "Plain GPU infrastructure API; no agent/harness tooling.", "checked": "2026-09-29" }, "ingress_rules": { "value": true, "source": "https://support.voltagepark.com/article/my-instances", "quote": "This section displays a list of port mappings that allow external access to services running on your instance.", "note": "VMs only: NAT port forwarding, you choose which internal ports are exposed via POST/DELETE /virtual-machines/{id}/port-forward (external port assigned from the host's available_ports). No IP allowlists or security groups. Bare metal gets a public IP with no documented provider-side filtering. The VAST storage tenant has a client-IP whitelist (storage access, not instance ingress).", "checked": "2026-09-29" }, "guest_firewall": { "value": true, "source": null, "quote": null, "note": "Full VMs and bare-metal servers with ubuntu user + sudo; own kernel.", "checked": "2026-09-29", "derived": true }, "secret_proxy": { "value": false, "source": "https://support.voltagepark.com/article/bare-metal", "quote": "Provide your Public SSH Key(s) > Select/Add organizational SSH public key(s) or expand the Advanced Configuration to add one-time public SSH public key(s).", "note": "Only SSH-key-based provisioning is documented; no secret-injection or outbound credential-proxy feature exists in the docs or API spec.", "checked": "2026-09-30" }, "volume_attach": { "value": true, "source": "https://support.voltagepark.com/article/attach-storage", "quote": "Your storage volumes can be attached to bare metal GPU instances for use. Each bare metal rental can be attached to a single storage volume.", "note": "VAST Data NFS 'shared storage' volumes (POST /storage/), resizable. Bare metal only: 'There is currently no support for attaching on-demand storage to a virtualized rental.' FAQ also says persistent storage is 'not enabled by default' (ask support).", "checked": "2026-09-29" }, "volume_shared": { "value": true, "source": "https://support.voltagepark.com/article/attach-storage", "quote": "A single storage volume can be attached to as many instances as you need.", "note": "Read-write NFSv3 mount (fstab example uses 'nfs rw'); root_squash/no_squash configurable via API. Bare metal (and its Managed Kubernetes clusters, which auto-mount) only.", "checked": "2026-09-29" }, "isolation": { "value": "qemu-kvm", "source": "https://support.voltagepark.com/article/introduction", "quote": "Virtual Machines (VMs) Optimized for development workflows that require fast startup, minimal latency, and easy customization. Bare Metal Built for customers that demand the highest performance and control... By removing virtualization layers, they provide direct access to hardware", "note": "The on-demand mode this card models is the VM tier (bare metal is a separate, non-virtualized offering). Docs confirm VMs are virtualized Linux machines with SSH as 'ubuntu' + full root, but never name the hypervisor; qemu-kvm is the standard choice for this VM shape on Linux GPU clouds. Derived, not confirmed by name.", "checked": "2026-09-30" }, "image_oci": { "value": false, "source": "https://support.voltagepark.com/article/virtual-machines-v-ms", "quote": "Configure your Virtual Machine by selecting the Number of GPUs > Continue. Provide your Public SSH Key(s)... A Deployment Name for your instance is optional.", "note": "Deployment flow only lets you pick GPU count, network, and SSH keys — no OS/image selection step and no custom Docker/OCI image upload documented anywhere in the On-Demand docs or OpenAPI spec.", "checked": "2026-09-30" }, "browser": { "value": false, "source": "https://www.voltagepark.com/pricing", "quote": "You can rent one H100 GPU for 1 hour starting at $1.99 without a contract", "note": "not offered: outside the product's scope — Voltage Park is raw GPU VM/bare-metal infrastructure, not a managed browser product.", "checked": "2026-09-30" }, "anti_bot": { "value": false, "source": "https://www.voltagepark.com/pricing", "quote": "You can rent one H100 GPU for 1 hour starting at $1.99 without a contract", "note": "not offered: outside the product's scope — GPU compute infrastructure, no browser/scraping stealth features.", "checked": "2026-09-30" }, "captcha_solving": { "value": false, "source": "https://www.voltagepark.com/pricing", "quote": "You can rent one H100 GPU for 1 hour starting at $1.99 without a contract", "note": "not offered: outside the product's scope — plain GPU cloud, no CAPTCHA-solving service.", "checked": "2026-09-30" }, "egress_ip_type": { "value": "datacenter", "source": "https://support.voltagepark.com/article/deploy-gp-us", "quote": "Public IP bandwidth 100 Gbps (Best Effort)", "note": "Standard datacenter public IPs assigned to GPU VM/bare-metal nodes; no residential/mobile IP product exists.", "checked": "2026-09-30" }, "pause_resume": { "value": true, "source": "https://support.voltagepark.com/article/my-instances", "quote": "Advanced > Deallocate compute resources: Release GPU, CPU, and RAM resources. Once complete, you will only be billed for storage.", "note": "Virtualized instances only: deallocating compute drops billing to storage-only while retaining the disk, and the instance can be restarted later (may relocate to a new host, taking up to 2 minutes). This matches 'pause and resume with reduced billing'. A plain Shut Down, by contrast, still bills at the full running rate.", "checked": "2026-09-30" }, "auto_stop_idle": { "value": false, "source": "https://support.voltagepark.com/article/my-instances", "quote": "Shut Down: Gracefully power off the instance while retaining its data. Stopped instances will still be charged at the full running billing rate.", "note": "not offered: only manual Shut Down / Terminate / Deallocate actions are documented; no idle-detection auto-stop is mentioned anywhere in the docs or OpenAPI spec.", "checked": "2026-09-30" }, "own_agent_api": { "value": false, "source": "https://support.voltagepark.com/article/api", "quote": "Our Voltage Park On Demand API allows developers to deploy, manage, and monitor virtual machines (VMs) and bare metal instances.", "note": "not offered: outside the product's scope — the API is plain GPU infrastructure management, not a first-party task/browser/computer-use agent.", "checked": "2026-09-30" }, "public_ipv4": { "value": true, "source": "https://support.voltagepark.com/article/virtual-machines-v-ms", "quote": "Use an SSH client to connect to your Virtual Machine (VM): ssh ubuntu@", "note": "Every instance (VM and bare metal) gets a public IPv4 address used for SSH/access.", "checked": "2026-09-30" }, "egress_allowlist": { "value": false, "source": "https://support.voltagepark.com/article/on-demand", "quote": "Clear Pricing Nothing hidden. Pay only for what you use. No hidden ingress, egress, or support costs.", "note": "not offered: docs and the full On-Demand OpenAPI spec describe only inbound port-forward mappings; no outbound allow/deny list, no-internet mode, or security-group egress control is documented.", "checked": "2026-09-30" }, "egress_http_rules": { "value": false, "source": "https://support.voltagepark.com/article/on-demand", "quote": "No hidden ingress, egress, or support costs.", "note": "not offered: no HTTP method/path-level egress controls are documented anywhere in the docs or OpenAPI spec.", "checked": "2026-09-30" }, "static_egress_ip": { "value": false, "source": "https://support.voltagepark.com/article/my-instances", "quote": "Instances in an \"only storage\" state may need to be relocated to a new host machine on the next startup. This can take up to 2 minutes, and can result in the IP address and external ports changing.", "note": "Docs explicitly say the instance's public IP can change across a deallocate/restart cycle, so the outbound IP is not guaranteed fixed; no static-IP product is offered.", "checked": "2026-09-30" }, "soc2": { "value": true, "source": "https://voltagepark.com/security", "quote": "SOC 2 Type II Certification", "note": null, "checked": "2026-09-30" }, "hipaa": { "value": true, "source": "https://voltagepark.com/security", "quote": "HIPAA Compliance", "note": null, "checked": "2026-09-30" }, "sso": { "value": false, "source": "https://support.voltagepark.com/article/account", "quote": "Organization Management This section is where you manage your team and invite new members to your organization.", "note": "not offered: account/organization docs only describe email-based team invites; no SAML/OIDC single sign-on is mentioned anywhere in the docs.", "checked": "2026-09-30" }, "eu_data_residency": { "value": false, "source": "https://www.voltagepark.com/pricing", "quote": "on-demand Ethernet 1-1016 HGX H100 GPUs", "note": "not offered: Voltage Park's on-demand GPU capacity is US-only (card regions: [\"us\"]); no EU region or data-residency option is documented.", "checked": "2026-09-30" }, "max_session_h": { "value": "none", "source": "https://www.voltagepark.com/pricing", "quote": "You can rent one H100 GPU for 1 hour starting at $1.99 without a contract", "note": "On-demand instances run until you stop/terminate them; no documented maximum session duration or forced timeout was found in the docs or OpenAPI spec.", "checked": "2026-09-30" }, "snapshot": { "value": "none", "source": "https://cloud-api.voltagepark.com/api/v1/openapi.json", "quote": "Shut Down: Gracefully power off the instance while retaining its data.", "note": "The On-Demand API (VMs, bare metal, storage, billing) has no snapshot/image endpoints. Instance lifecycle is only Shut Down (keeps disk, still billed) or Terminate (data deleted) plus 'Deallocate compute resources' (storage-only billing). No machine-image/snapshot concept anywhere in the help center or API spec.", "checked": "2026-09-30" }, "fork": { "value": false, "source": "https://cloud-api.voltagepark.com/api/v1/openapi.json", "quote": "Terminate: Permanently delete the instance and its associated data.", "note": "No snapshot/image capability exists to fork from (see snapshot), and no clone/duplicate-instance endpoint or docs mention exists. Instances are only created fresh, shut down, or terminated.", "checked": "2026-09-30" }, "fork_running": { "value": false, "source": "https://cloud-api.voltagepark.com/api/v1/openapi.json", "quote": "Terminate: Permanently delete the instance and its associated data.", "note": "No live-instance forking/branching capability documented or exposed via the On-Demand API; consistent with no snapshot support at all.", "checked": "2026-09-30" }, "memory_snapshot_fork": { "value": false, "source": "https://cloud-api.voltagepark.com/api/v1/openapi.json", "quote": "Terminate: Permanently delete the instance and its associated data.", "note": "No memory/VM-state snapshot mechanism exists at all (only disk persists across Shut Down); no basis for memory-snapshot forking.", "checked": "2026-09-30" }, "desktop": { "value": false, "source": "https://support.voltagepark.com/article/bare-metal", "quote": "Use an SSH client to connect to your Bare Metal instance: ssh ubuntu@", "note": "Access is SSH-only to headless Ubuntu Linux GPU servers/VMs; no browser desktop / VNC / remote-GUI product exists. Product is a raw GPU cloud, not a desktop environment.", "checked": "2026-09-30" }, "computer_use": { "value": false, "source": "https://support.voltagepark.com/article/on-demand", "quote": "Get instant access to bare metal NVIDIA HGX H100 with no long-term commitment, paying only for what you use at budget-friendly prices.", "note": "Plain GPU compute rental (SSH/bare-metal access); no computer-use / GUI-automation product offered.", "checked": "2026-09-30" }, "wake_on_request": { "value": false, "source": "https://support.voltagepark.com/article/my-instances", "quote": "Please note that there is an approximately 20-minute delay when restarting instances after they have been stopped.", "note": "Instances are manually started/stopped and take ~20 minutes to restart; no request-triggered autoscale/wake-on-request mechanism documented.", "checked": "2026-09-30" }, "inbound_https": { "value": false, "source": "https://support.voltagepark.com/article/my-instances", "quote": "Forwarded Ports (Virtual Machines Only): This section displays a list of port mappings that allow external access to services running on your instance.", "note": "Only raw NAT port-forwarding (VMs) or a plain public IP (bare metal) is offered; no managed HTTPS/TLS-terminated ingress or reverse proxy.", "checked": "2026-09-30" }, "agent_harnesses": { "value": false, "source": "https://support.voltagepark.com/article/api", "quote": "Our Voltage Park On Demand API allows developers to deploy, manage, and monitor virtual machines (VMs) and bare metal instances.", "note": "Plain GPU infrastructure API for provisioning VMs/bare metal; no agent-harness / agent-framework integration documented.", "checked": "2026-09-30" }, "mcp_server": { "value": false, "source": "https://support.voltagepark.com/article/api", "quote": "Our Voltage Park On Demand API allows developers to deploy, manage, and monitor virtual machines (VMs) and bare metal instances.", "note": "The only integration surface is a REST API for VM/bare-metal management; no MCP server product or docs exist.", "checked": "2026-09-30" }, "windows": { "value": false, "source": "https://support.voltagepark.com/article/bare-metal", "quote": "Use an SSH client to connect to your Bare Metal instance: ssh ubuntu@", "note": "Instances log in as the 'ubuntu' user via SSH; product is Linux (Ubuntu) only, no Windows image/OS offered.", "checked": "2026-09-30" }, "macos": { "value": false, "source": "https://support.voltagepark.com/article/bare-metal", "quote": "Voltage Park offers Bare Metal as the ultimate choice for performance, control, and customization. By removing virtualization layers, they provide direct access to hardware for maximum efficiency.", "note": "Product is NVIDIA HGX H100 datacenter GPU servers running Linux; macOS is not a supported guest OS and is outside what this GPU cloud offers.", "checked": "2026-09-30" }, "arm64": { "value": false, "source": "https://www.voltagepark.com/pricing", "quote": "NVIDIA H100 Pricing Transparent, competitive pricing that works within your budget.", "note": "Voltage Park sells HGX H100 GPU servers (x86 host CPUs paired with NVIDIA H100s); no ARM64 instance type is advertised anywhere on the pricing or docs pages.", "checked": "2026-09-30" }, "self_host": { "value": false, "source": "https://voltagepark.com/", "quote": "Voltage Park provides on-demand and reserved GPU cloud infrastructure. Scalable secure computing power with latest NVIDIA HGX H100 GPUs and cutting-edge data centers.", "note": "Voltage Park is a hosted GPU-cloud service run in its own datacenters; there is no self-hosted/on-prem software distribution of its platform.", "checked": "2026-09-30" }, "byoc": { "value": false, "source": "https://voltagepark.com/", "quote": "Voltage Park provides on-demand and reserved GPU cloud infrastructure.", "note": "Capacity is rented in Voltage Park's own data centers; no bring-your-own-cloud/account option is offered or documented.", "checked": "2026-09-30" }, "open_source": { "value": false, "source": "https://voltagepark.com/", "quote": "Voltage Park provides on-demand and reserved GPU cloud infrastructure.", "note": "Commercial closed-source cloud platform; no open-source release of the platform itself is documented.", "checked": "2026-09-30" }, "live_resize": { "value": false, "source": "https://support.voltagepark.com/article/bare-metal", "quote": "Configure your Bare Metal instance by selecting the Number of GPUs (in multiples of 8x GPUs) > Continue.", "note": "GPU/CPU/RAM counts are fixed at deployment time via the deploy wizard; no documented way to live-resize a running instance's resources (only Shut Down/Terminate/Deallocate lifecycle actions exist).", "checked": "2026-09-30" }, "nested_virt": { "value": false, "source": "https://support.voltagepark.com/article/virtual-machines-v-ms", "quote": "Voltage Park's Virtual Machines (VMs) are built with state-of-the-art hardware and optimized for small projects.", "note": "not offered: the on-demand mode this card models is the standard virtualized VM tier (qemu-kvm); no docs, deploy wizard option, or OpenAPI field mentions /dev/kvm, nested virtualization, or exposing hardware virtualization extensions inside the guest. Bare metal (a separate, non-virtualized tier) would have real hardware KVM support, but that isn't the mode this card models.", "checked": "2026-09-30" }, "lang_node": { "value": false, "source": "https://support.voltagepark.com/article/virtual-machines-v-ms", "quote": "Provide your Public SSH Key(s)... A Deployment Name for your instance is optional... Use an SSH client to connect to your Virtual Machine (VM): ssh ubuntu@", "note": "Deploy flow only configures GPU count and SSH keys; there is no documented base-image software list, template, or first-party Node.js SDK/code-exec. Instances are plain Ubuntu VMs reached via SSH as root/ubuntu with no preinstalled runtime stack described anywhere in the docs or API spec, so Node.js is not available out of the box (users install it themselves via sudo).", "checked": "2026-09-30" }, "usage_api": { "value": true, "source": "https://cloud-api.voltagepark.com/api/v1/openapi.json", "quote": "GET /billing/transactions returns HistoricalTransactionInTimePeriod objects with linked_instance.virtual_machine_id for each VM charge, including total_amount in dollars", "note": "granularity: per-sandbox; cost: yes; Voltage Park's On-Demand API exposes /billing/transactions endpoint which returns paginated historical billing records. Each transaction record for VM usage is linked to a specific virtual_machine_id (UUID) and includes total_amount (cost). Transactions can be filtered by type (including 'virtual_machine') and by date range (earliest, latest parameters). This allows per-VM/sandbox billing tracking.", "confidence": "high", "checked": "2026-10-01" } }, "locations": [ { "country": "US", "state": "TX", "city": "Dallas", "provider_region": "dallas", "self_serve": true, "choosable": true } ], "locations_source": "https://cloud-api.voltagepark.com/api/v1/bare-metal/locations", "locations_quote": "H100 GPU rental starting at $1.99/hour with on-demand Ethernet access in Dallas.", "locations_checked": "2026-10-01" }