{
 "features_resolved": {
  "snapshot_any": true,
  "snapshot_mem": true,
  "fork": true,
  "pause_resume": true,
  "persistent_disk": true,
  "whole_disk": null,
  "persist_scope": null,
  "volumes": false,
  "auto_stop_idle": true,
  "long_sessions": true,
  "custom_image": true,
  "image_snapshot": true,
  "image_oci": false,
  "vm_isolation": true,
  "docker_inside": true,
  "nested_virt": null,
  "root": true,
  "systemd": null,
  "browser": true,
  "desktop": true,
  "computer_use": true,
  "code_interpreter": null,
  "browser_control": true,
  "desktop_control": true,
  "browser_and_desktop": true,
  "anti_bot": false,
  "captcha_solving": false,
  "residential_ip": true,
  "agent_harnesses": true,
  "ssh": true,
  "public_ipv4": false,
  "inbound_https": true,
  "custom_domain": true,
  "raw_tcp_inbound": null,
  "egress_allowlist": true,
  "open_internet": true,
  "static_egress_ip": true,
  "private_network": null,
  "self_host": true,
  "byoc": true,
  "open_source": false,
  "soc2": null,
  "hipaa": null,
  "sso": null,
  "gpu": true,
  "arm64": false,
  "windows": true,
  "macos": true,
  "lang_python": true,
  "lang_node": true,
  "lang_go": null,
  "lang_rust": null,
  "lang_java": null,
  "gpu_desktop": true,
  "credential_injection": true,
  "egress_http_rules": false,
  "wake_on_request": true,
  "live_resize": true,
  "fork_running": true,
  "memory_snapshot_fork": true,
  "webhooks": true,
  "scheduled_wakeups": null,
  "audit_logs": null,
  "eu_data_residency": false,
  "zero_data_retention": null,
  "scoped_api_keys": true,
  "spend_limits": true,
  "mcp_server": true,
  "usage_api": false,
  "snapshot_auto": false,
  "snapshot_on_demand": true,
  "harness_api": false,
  "own_agent_api": true,
  "fast_boot": null,
  "strong_isolation": true,
  "ingress_rules": true,
  "guest_firewall": true,
  "secret_proxy": false,
  "secret_proxy_any": false,
  "volume_attach": false,
  "volume_shared": false,
  "sidecar": null,
  "fuse": true,
  "kernel_features": true,
  "burst": null
 },
 "company": {
  "id": "orgo",
  "checked": "2026-09-30",
  "company": "Organic Intelligence, Inc. (Orgo)",
  "ownership": {
   "value": "independent",
   "parent": null,
   "source": "https://www.orgo.ai/",
   "quote": "Backed by Y Combinator"
  },
  "stage": {
   "value": "pre-seed",
   "source": "https://www.orgo.ai/",
   "quote": "Backed by Y Combinator",
   "confidence": "medium",
   "note": "No priced round found. Co-founder's LinkedIn lists Orgo as YC F26 (Fall 2026 batch, current at time of check), implying only the (undisclosed-amount) YC standard deal is confirmed; mapped to the closest stage, pre-seed."
  },
  "accelerators": [
   {
    "name": "Y Combinator",
    "batch": "F26",
    "source": "https://duckduckgo.com/html/?q=%22Tiger+Wang%22+Orgo+Y+Combinator",
    "quote": "AI Virtual Desktop | Co-founder @ Orgo (YC F26) · Experience: Orgo · Education: University of Illinois Urbana-Champaign · Location: San Francisco",
    "confidence": "medium",
    "via": "research"
   }
  ],
  "total_funding_usd": {
   "value": null,
   "status": "unknown"
  },
  "last_round": {
   "type": "unknown",
   "amount_usd": null,
   "date": null,
   "leads": [],
   "status": "unknown",
   "source": "",
   "quote": ""
  },
  "investors": [
   "Y Combinator"
  ],
  "portfolio_listings": [],
  "revenue": {
   "value": null,
   "status": "unknown"
  },
  "employees": {
   "value": 3,
   "source": "https://www.linkedin.com/company/orgoai"
  },
  "note": "Orgo (legal name Organic Intelligence, Inc.) sells cloud/VM 'computers for AI agents' (persistent Linux/Windows desktops, sub-500ms boot claims). Founded 2023, San Francisco; co-founders Tiger Wang and Nick Vasilescu. Self-described 'Backed by Y Combinator' on its homepage, and a co-founder's LinkedIn lists the YC batch as F26 (Fall 2026). No priced funding round, total funding amount, or revenue figures were found in public sources within the research window — buyers should treat funding/revenue as unverified. Caution: the domain/brand 'orgo.ai' is shared in search results with at least two unrelated companies (a hospital OR-staffing platform and an older networking/telecom firm) — do not conflate their Crunchbase/Getlatka/PrivCo profiles with this Orgo."
 },
 "id": "orgo",
 "name": "Orgo",
 "url": "https://www.orgo.ai/pricing",
 "category": "agent-sandbox",
 "isolation": "vm",
 "modes": [
  {
   "key": "pool",
   "label": "Monthly plan pool (Linux)",
   "flags": [],
   "os": [
    "linux"
   ],
   "gpu_only": false,
   "features": {
    "isolation": "vm"
   },
   "pricing": "pool",
   "vcpu_h": null,
   "ram_gib_h": null,
   "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": 4,
   "max_ram_gib": 64,
   "sizes": null,
   "pool_tiers": [
    {
     "name": "Starter (1 computer, 8 GB RAM, 40 GB disk)",
     "month": 29,
     "vcpu": null,
     "ram_gib": 8,
     "max_vms": 1
    },
    {
     "name": "Growth (4 computers, 32 GB RAM pool, 160 GB)",
     "month": 99,
     "vcpu": 16,
     "ram_gib": 32,
     "max_vms": 4
    },
    {
     "name": "Growth + 1 extra Medium (5 computers)",
     "month": 131.5,
     "vcpu": 18,
     "ram_gib": 40,
     "max_vms": 5
    },
    {
     "name": "Growth + 6 extra Medium (10 computers)",
     "month": 294,
     "vcpu": 28,
     "ram_gib": 80,
     "max_vms": 10
    },
    {
     "name": "Scale (16 computers, 128 GB RAM pool, 640 GB)",
     "month": 399,
     "vcpu": 64,
     "ram_gib": 128,
     "max_vms": 16
    },
    {
     "name": "Scale + 9 extra Medium (25 computers)",
     "month": 691.5,
     "vcpu": 82,
     "ram_gib": 200,
     "max_vms": 25
    },
    {
     "name": "Scale + 34 extra Medium (50 computers)",
     "month": 1504,
     "vcpu": 132,
     "ram_gib": 400,
     "max_vms": 50
    }
   ],
   "requires_always_on": false,
   "always_on_month_per_instance": null,
   "commit_note": null,
   "min_billed_seconds": 0,
   "granularity_s": 1,
   "start_fee": 0,
   "boot_overhead_s": null,
   "multiplier": 1,
   "windows_vcpu_h": null,
   "gpu": null,
   "note": "Flat monthly plan = N always-on computers with an account RAM pool (8 GB per included computer). Computers \"stay running until you stop or delete them\"; no per-hour charge on plans. Per-computer ceiling without add-ons: Starter 8 GB RAM / 40 GB disk, Growth 16 GB / 80 GB, Scale 32 GB / 150 GB; hard cap 4 vCPU / 64 GB / 300 GB on any plan (pricing.md). Pool vCPU is NOT published: encoded as included computers x 4 vCPU (the hard per-computer cap), an upper bound, plus 2 vCPU per extra Medium. Extra-computer rows are the cheapest setups pricing.md publishes for 5/10/25/50 computers (extra Medium $32.50/mo = 2 vCPU / 8 GB / 40 GB); counts between them price at the next published setup, an upper bound. Plan AI credit ($5/$10/$50 per month) is model-usage credit, not compute credit."
  },
  {
   "key": "pool-windows",
   "label": "Monthly plan pool (Windows)",
   "flags": [],
   "os": [
    "windows"
   ],
   "gpu_only": false,
   "features": {
    "isolation": "vm"
   },
   "pricing": "pool",
   "vcpu_h": null,
   "ram_gib_h": null,
   "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": 4,
   "max_ram_gib": 64,
   "sizes": null,
   "pool_tiers": [
    {
     "name": "Starter + 1 Windows licence (1 computer, 8 GB RAM)",
     "month": 44,
     "vcpu": null,
     "ram_gib": 8,
     "max_vms": 1
    },
    {
     "name": "Growth + 4 Windows licences (4 computers, 32 GB RAM pool)",
     "month": 159,
     "vcpu": 16,
     "ram_gib": 32,
     "max_vms": 4
    },
    {
     "name": "Scale (16 computers, 128 GB RAM pool, 640 GB; Windows included)",
     "month": 399,
     "vcpu": 64,
     "ram_gib": 128,
     "max_vms": 16
    }
   ],
   "requires_always_on": false,
   "always_on_month_per_instance": null,
   "commit_note": null,
   "min_billed_seconds": 0,
   "granularity_s": 1,
   "start_fee": 0,
   "boot_overhead_s": null,
   "multiplier": 1,
   "windows_vcpu_h": null,
   "gpu": null,
   "note": "Windows on Scale (included) or on Starter/Growth with a $15/mo Windows licence per Windows computer (\"The license is a COUNT, not a label on a particular computer\"). Windows needs 8 GB RAM or more. Starter/Growth rows = plan + one licence per included computer. Flat monthly plan = N always-on computers with an account RAM pool (8 GB per included computer). Computers \"stay running until you stop or delete them\"; no per-hour charge on plans. Per-computer ceiling without add-ons: Starter 8 GB RAM / 40 GB disk, Growth 16 GB / 80 GB, Scale 32 GB / 150 GB; hard cap 4 vCPU / 64 GB / 300 GB on any plan (pricing.md). Pool vCPU is NOT published: encoded as included computers x 4 vCPU (the hard per-computer cap), an upper bound, plus 2 vCPU per extra Medium. Plan AI credit ($5/$10/$50 per month) is model-usage credit, not compute credit."
  },
  {
   "key": "reserved-1y",
   "label": "Yearly billing (Linux)",
   "flags": [
    "sales"
   ],
   "os": [
    "linux"
   ],
   "gpu_only": false,
   "features": {
    "isolation": "vm"
   },
   "pricing": "pool",
   "vcpu_h": null,
   "ram_gib_h": null,
   "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": 4,
   "max_ram_gib": 64,
   "sizes": null,
   "pool_tiers": [
    {
     "name": "Starter (1 computer, 8 GB RAM, 40 GB disk)",
     "month": 24.17,
     "vcpu": null,
     "ram_gib": 8,
     "max_vms": 1
    },
    {
     "name": "Growth (4 computers, 32 GB RAM pool, 160 GB)",
     "month": 82.5,
     "vcpu": 16,
     "ram_gib": 32,
     "max_vms": 4
    },
    {
     "name": "Growth + 1 extra Medium (5 computers)",
     "month": 109.58,
     "vcpu": 18,
     "ram_gib": 40,
     "max_vms": 5
    },
    {
     "name": "Growth + 6 extra Medium (10 computers)",
     "month": 245,
     "vcpu": 28,
     "ram_gib": 80,
     "max_vms": 10
    },
    {
     "name": "Scale (16 computers, 128 GB RAM pool, 640 GB)",
     "month": 332.5,
     "vcpu": 64,
     "ram_gib": 128,
     "max_vms": 16
    },
    {
     "name": "Scale + 9 extra Medium (25 computers)",
     "month": 576.25,
     "vcpu": 82,
     "ram_gib": 200,
     "max_vms": 25
    },
    {
     "name": "Scale + 34 extra Medium (50 computers)",
     "month": 1253.33,
     "vcpu": 132,
     "ram_gib": 400,
     "max_vms": 50
    }
   ],
   "requires_always_on": false,
   "always_on_month_per_instance": null,
   "commit_note": "Billed yearly up front: Starter $290/yr, Growth $990/yr, Scale $3,990/yr (month = /12). Yearly extra Medium $27.08/mo ($325/yr), bought through the capacity API (\"term\": \"yearly\"); setup rows use pricing.md's \"Yearly, per month\" column.",
   "min_billed_seconds": 0,
   "granularity_s": 1,
   "start_fee": 0,
   "boot_overhead_s": null,
   "multiplier": 1,
   "windows_vcpu_h": null,
   "gpu": null,
   "note": "Flat monthly plan = N always-on computers with an account RAM pool (8 GB per included computer). Computers \"stay running until you stop or delete them\"; no per-hour charge on plans. Per-computer ceiling without add-ons: Starter 8 GB RAM / 40 GB disk, Growth 16 GB / 80 GB, Scale 32 GB / 150 GB; hard cap 4 vCPU / 64 GB / 300 GB on any plan (pricing.md). Pool vCPU is NOT published: encoded as included computers x 4 vCPU (the hard per-computer cap), an upper bound, plus 2 vCPU per extra Medium. Extra-computer rows are the cheapest setups pricing.md publishes for 5/10/25/50 computers (extra Medium $32.50/mo = 2 vCPU / 8 GB / 40 GB); counts between them price at the next published setup, an upper bound. Plan AI credit ($5/$10/$50 per month) is model-usage credit, not compute credit."
  },
  {
   "key": "reserved-1y-windows",
   "label": "Yearly billing (Windows)",
   "flags": [
    "sales"
   ],
   "os": [
    "windows"
   ],
   "gpu_only": false,
   "features": {
    "isolation": "vm"
   },
   "pricing": "pool",
   "vcpu_h": null,
   "ram_gib_h": null,
   "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": 4,
   "max_ram_gib": 64,
   "sizes": null,
   "pool_tiers": [
    {
     "name": "Starter + 1 Windows licence (1 computer, 8 GB RAM)",
     "month": 36.67,
     "vcpu": null,
     "ram_gib": 8,
     "max_vms": 1
    },
    {
     "name": "Growth + 4 Windows licences (4 computers, 32 GB RAM pool)",
     "month": 132.5,
     "vcpu": 16,
     "ram_gib": 32,
     "max_vms": 4
    },
    {
     "name": "Scale (16 computers, 128 GB RAM pool, 640 GB; Windows included)",
     "month": 332.5,
     "vcpu": 64,
     "ram_gib": 128,
     "max_vms": 16
    }
   ],
   "requires_always_on": false,
   "always_on_month_per_instance": null,
   "commit_note": "Billed yearly up front: Starter $290/yr, Growth $990/yr, Scale $3,990/yr; Windows licence $150/yr ($12.50/mo) per Windows computer on Starter/Growth, included on Scale.",
   "min_billed_seconds": 0,
   "granularity_s": 1,
   "start_fee": 0,
   "boot_overhead_s": null,
   "multiplier": 1,
   "windows_vcpu_h": null,
   "gpu": null,
   "note": "Windows on Scale (included) or on Starter/Growth with a $15/mo Windows licence per Windows computer (\"The license is a COUNT, not a label on a particular computer\"). Windows needs 8 GB RAM or more. Starter/Growth rows = plan + one licence per included computer. Flat monthly plan = N always-on computers with an account RAM pool (8 GB per included computer). Computers \"stay running until you stop or delete them\"; no per-hour charge on plans. Per-computer ceiling without add-ons: Starter 8 GB RAM / 40 GB disk, Growth 16 GB / 80 GB, Scale 32 GB / 150 GB; hard cap 4 vCPU / 64 GB / 300 GB on any plan (pricing.md). Pool vCPU is NOT published: encoded as included computers x 4 vCPU (the hard per-computer cap), an upper bound, plus 2 vCPU per extra Medium. Plan AI credit ($5/$10/$50 per month) is model-usage credit, not compute credit."
  },
  {
   "key": "enterprise-mac",
   "label": "Enterprise / by request (50+ computers; Mac and iPhone)",
   "flags": [
    "sales",
    "beta"
   ],
   "os": [
    "macos",
    "windows",
    "linux"
   ],
   "gpu_only": false,
   "features": {
    "isolation": "vm"
   },
   "pricing": "pool",
   "vcpu_h": null,
   "ram_gib_h": null,
   "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": 0,
   "granularity_s": 1,
   "start_fee": 0,
   "boot_overhead_s": null,
   "multiplier": 1,
   "windows_vcpu_h": null,
   "gpu": null,
   "note": "\"Custom. Volume pricing, billed annually.\" \"50+ computers Up to 64 GB RAM per computer\". Mac and iPhone are by request (Enterprise, or an approved request from Settings): \"A Mac (a macOS computer on a Mac mini) starts at $90/mo\" (M4 VM $100/mo, M6 VM $120/mo), \"a physical iPhone at $65/mo\"; the account price is set on approval, so not priced here. macOS RAM is fixed at 10 GB. Shared, dedicated or on-prem."
  },
  {
   "key": "enterprise-gpu",
   "label": "Enterprise GPU/hardware fleet",
   "flags": [
    "sales"
   ],
   "os": [
    "linux",
    "windows"
   ],
   "gpu_only": true,
   "pricing": "pool",
   "pool_tiers": null,
   "gpu": {
    "T4": null,
    "L4": null,
    "A10": null,
    "L40S": null,
    "A100-40GB": null,
    "A100-80GB": null,
    "H100": null
   },
   "note": "Custom annual/volume pricing; no public GPU price or availability SLA. The T4..H100 list comes from the template `gpu` field, which the template schema now marks \"Accepted, but not used at launch\", so it is not evidence those GPUs launch. Self-serve vGPU slices (2q / 4q) exist on Linux create but are unpriced (see caveats)."
  }
 ],
 "storage": {
  "included_disk_gib": null,
  "disk_gib_month": null,
  "disk_billed_when_stopped": true,
  "snapshot_gib_month": null,
  "paused_ram_billed": false,
  "daily_backups": "Daily filesystem snapshot backups for every computer, with restore in Settings; old snapshots are aged out but the public retention duration and marginal price are not stated."
 },
 "network": {
  "egress_free_gib": null,
  "egress_gib": null,
  "ipv4_month": null,
  "ipv4_available": false,
  "internet": {
   "default": "full"
  }
 },
 "plans": [
  {
   "name": "Self-serve (fee carried in pool tiers)",
   "fee": 0,
   "flags": [],
   "fee_is_credit": false,
   "included_usd": 0,
   "per_seat": 0,
   "concurrency": null,
   "max_session_h": null,
   "max_vcpu": null,
   "max_ram_gib": null,
   "sizes_allowed": null,
   "note": "Starter 1 seat, Growth 5 seats, Scale 25 seats. Extra computers are add-ons on any plan, per computer per month: Small $14 (API only), Medium $32.50, Large $64.50, XL $108.50, 2XL $195.50; self-serve up to +100 computers, raised on request \"at the same prices\"."
  },
  {
   "name": "Enterprise",
   "fee": null,
   "flags": [
    "sales"
   ],
   "fee_is_credit": false,
   "included_usd": 0,
   "per_seat": 0,
   "concurrency": null,
   "max_session_h": null,
   "max_vcpu": null,
   "max_ram_gib": null,
   "sizes_allowed": null,
   "note": "50+ computers, up to 64 GB RAM per computer, 50+ team members, Linux + Windows + Mac, shared/dedicated/on-prem, SLA."
  },
  {
   "name": "Legacy",
   "fee": null,
   "flags": [
    "legacy"
   ],
   "fee_is_credit": false,
   "included_usd": 0,
   "per_seat": null,
   "concurrency": null,
   "max_session_h": null,
   "note": "Pre-2026-07-06 subscribers keep their old named plan and limits; switching cost is shown in account."
  }
 ],
 "free": {
  "monthly_credit": 0,
  "one_time_credit": 0
 },
 "features": {
  "linux": true,
  "windows": true,
  "macos": true,
  "arm64": false,
  "gpu": true,
  "gpu_types": [],
  "snapshot": "mem",
  "pause_resume": true,
  "fork": true,
  "persistent_disk": true,
  "volumes": null,
  "auto_stop_idle": true,
  "max_session_h": null,
  "custom_image": true,
  "docker_inside": true,
  "nested_virt": null,
  "root": true,
  "systemd": null,
  "browser": true,
  "desktop": true,
  "computer_use": true,
  "code_interpreter": null,
  "agent_harnesses": true,
  "ssh": true,
  "public_ipv4": false,
  "inbound_https": true,
  "custom_domain": true,
  "raw_tcp_inbound": null,
  "egress_allowlist": true,
  "static_egress_ip": true,
  "private_network": null,
  "self_host": true,
  "byoc": true,
  "open_source": false,
  "regions": [
   "us"
  ],
  "soc2": null,
  "hipaa": null,
  "sso": null,
  "lang_python": true,
  "lang_node": true,
  "lang_go": null,
  "lang_rust": null,
  "lang_java": null,
  "gpu_desktop": true,
  "credential_injection": true,
  "egress_http_rules": false,
  "wake_on_request": true,
  "live_resize": true,
  "fork_running": true,
  "memory_snapshot_fork": true,
  "webhooks": true,
  "scheduled_wakeups": null,
  "audit_logs": null,
  "eu_data_residency": false,
  "zero_data_retention": null,
  "scoped_api_keys": true,
  "spend_limits": true,
  "mcp_server": true,
  "browser_control": true,
  "desktop_control": true,
  "anti_bot": false,
  "captcha_solving": false,
  "egress_ip_type": "datacenter; residential beta by allowlist; customer-supplied HTTP CONNECT proxy supported",
  "ip_note": "Computers egress from data-center IPs; BYO upstream HTTP CONNECT proxy for browser traffic via /egress/upstream (changes exit IP for browser only); no managed residential proxies documented.",
  "isolation": "vm",
  "image_oci": false,
  "android": true,
  "snapshot_auto": false,
  "snapshot_on_demand": true,
  "own_agent_api": true,
  "harness_api": false,
  "egress_allowlist_level": "both",
  "ingress_rules": true,
  "guest_firewall": true,
  "secret_proxy": false,
  "volume_attach": false,
  "volume_shared": false,
  "max_session_unlimited": true,
  "open_internet": true,
  "usage_api": false
 },
 "perf": {
  "cpu_runs_s": null,
  "cold_start_ms": null,
  "burst_tti_ms": null,
  "dax_s": null
 },
 "knobs": [],
 "caveats": [
  "Plans renamed (live pricing 2026-10-14): Hacker -> Starter ($29, \"1 computer 8 GB RAM, 40 GB disk\"), Startup -> Growth ($99, \"4 computers Up to 16 GB RAM per computer 32 GB RAM in total\"), Scale ($399, \"16 computers Up to 32 GB RAM per computer 128 GB RAM in total\"), Enterprise \"50+ computers Up to 64 GB RAM per computer\". Starter vCPU is not published, so its pool vCPU is null and it is not auto-picked (an 8 GB computer is the Medium 2 vCPU shape).",
  "Per-computer RAM/disk ceilings differ by plan (Starter 8 GB / 40 GB, Growth 16 GB / 80 GB, Scale 32 GB / 150 GB without add-ons; 64 GB / 300 GB with \"Room to grow\" add-ons). The engine has no per-tier per-computer ceiling, so e.g. one 32 GB computer may be priced on Growth, which would need RAM add-ons. Logged in .work/engine-needs/orgo.md.",
  "Extra computers: \"Plans start at $29 a month. Add computers as you grow, at $32.50 each.\" (live pricing). The engine cannot add N extra computers to a tier, so the card carries the cheapest setups pricing.md publishes (Growth + 1 / + 6, Scale + 9 / + 34 extra Medium for 5 / 10 / 25 / 50 computers). Fleet sizes between those price at the next published setup (an upper bound), and workloads needing more than 2 vCPU or 8 GB per extra computer (Large $64.50, XL $108.50, 2XL $195.50, or vCPU add-ons) are not modelled. Self-serve add-on cap +100 computers; above that the cap is raised on request at the same prices; Enterprise prices from 50 computers by volume.",
  "Windows: on Scale (included) or on any paid plan with a $15/mo licence per Windows computer ($150/yr); Windows needs 8 GB RAM or more. Scale + extra-computer setups are not modelled for Windows (whether extras on Scale need a licence is not stated). macOS (Mac from $90/mo) and physical iPhones ($65/mo) only by request; Android is an accepted OS value where a host supports it.",
  "Self-serve GPU: Linux computers can attach a vGPU slice, \"`2q` (2 GB VRAM) or `4q` (4 GB VRAM)\" (create docs), but GPU model and price are unpublished and creates may fall back to CPU when no GPU host is free, so GPU workloads are not priced.",
  "Hourly billing also exists (create docs: `billing_mode` \"`hourly` runs it on credits by the hour\"; Medium Linux only, no GPU or custom template): pricing.md quotes \"$0.06/hr while it runs and $0.004/hr while it is stopped\" for the free tier's Medium. Its availability beyond the $5 signup credit and per-account limits (agent accounts: up to 5 hourly Medium computers) are not stated clearly enough to model as a mode.",
  "Locations: the template `region` field is \"Accepted, but not used for placement\" and the CLI `--region` is only a hint, so the US cities are not choosable.",
  "Pool vCPU per tier not published; set to computers x 4 vCPU (per-computer hard cap), so large-vCPU workloads may be accepted that the plan would refuse (403 quota).",
  "RAM pool per plan: 8 / 32 / 128 GB (Starter / Growth / Scale), 8 GB per included computer; each extra Medium adds 8 GB. A 4 vCPU / 8 GB workload fits 1 / 4 / 16 computers on the base plans.",
  "Stopping a computer archives its disk (new host + new IP on start, no memory state). Idle time is not billed separately because billing is flat.",
  "Monthly price verified on live page; yearly prices read after toggling \"Yearly\": $24/$82/$332 per month, \"$290 / $990 / $3,990 billed yearly\".",
  "Category changed windows -> agent-sandbox (Linux is the default OS on all paid tiers). Yearly modes flagged sales (annual prepay, opt-in). Retained computer slots count ALL existing computers (stopped ones too) until deleted, not only running peak; stopping does not reduce the fee. Per-computer CPU values accepted: 0.5/1/2/4 (6/8/16 pass validation then 403 PER_COMPUTER_CPU_CAP); RAM 4/8/16/32/64 GB; API default 1 CPU/4 GB, dashboard default Medium 2 CPU/8 GB. (patch v3-computer-use, 2026-09-28)"
 ],
 "sources": [
  "https://www.orgo.ai/pricing",
  "https://docs.orgo.ai/guides/instance-types",
  "https://docs.orgo.ai/llms.txt",
  "https://docs.orgo.ai/llms-full.txt",
  "https://docs.orgo.ai/guides/mcp",
  "https://www.orgo.ai/pricing.md",
  "https://docs.orgo.ai/api-reference/computers/create",
  "https://docs.orgo.ai/guides/templates/schema",
  "https://docs.orgo.ai/api-reference/account/add-capacity",
  "https://docs.orgo.ai/guides/openclaw"
 ],
 "fleet": {
  "default_max_instances": 1,
  "default_max_vcpu": null,
  "default_max_gpus": null,
  "raise": "self-serve",
  "raise_time_days": null,
  "verification": null,
  "tiers": "Starter 1 computer; Growth 4; Scale 16; +100 extra computers self-serve on any plan ($32.50/mo per Medium); above that the cap is raised on request at the same prices; Enterprise 50+ by volume.",
  "notes": "Entry paid count 1, not a demonstrated free-trial fleet cap. 10 = Growth + 6 extra ($294), 50 = Scale + 34 extra ($1,504) per pricing.md; self-serve add-on cap +100 computers / +800 GB RAM / +4 TB disk (vCPU cap stated as +200 on pricing.md, 50 in the add-capacity docs). 300/1,000 need a cap raise on request (\"to thousands of computers if you need them\"). Per-computer and account RAM caps additional.",
  "sources": [
   "https://www.orgo.ai/pricing",
   "https://www.orgo.ai/pricing.md",
   "https://docs.orgo.ai/api-reference/account/add-capacity",
   "https://docs.orgo.ai/guides/instance-types"
  ],
  "researched": "2026-10-14"
 },
 "audit_2026_09_29": {
  "applied": 22,
  "source": "checked against the provider's live pricing and docs pages",
  "reviewed": 2
 },
 "extras": [
  {
   "item": "Extra computer (Medium: 8 GB, 2 vCPU, 40 GB)",
   "price": "$32.50 per computer per month ($27.08/mo yearly); Small $14, Large $64.50, XL $108.50, 2XL $195.50"
  },
  {
   "item": "RAM component",
   "price": "$2.50 per GB per month; per-computer hard cap 64 GB"
  },
  {
   "item": "vCPU component",
   "price": "$2.50 per 0.5 vCPU per month ($5/vCPU-month); per-computer hard cap 4 vCPU"
  },
  {
   "item": "Disk component",
   "price": "$1 per 20 GB block per month; per-computer hard cap 300 GB"
  },
  {
   "item": "Computer slot component",
   "price": "$0.50 per extra computer per month"
  },
  {
   "item": "Windows licence",
   "price": "$15 per Windows computer per month on Starter/Growth; included on Scale"
  },
  {
   "item": "Mac / iPhone (by request)",
   "price": "Mac from $90/mo (M4 VM $100, M6 VM $120), physical iPhone $65/mo; set per account on approval"
  },
  {
   "item": "vGPU slice (2q / 4q, Linux)",
   "price": "not published"
  }
 ],
 "feature_evidence": {
  "snapshot_auto": {
   "value": false,
   "source": "https://docs.orgo.ai/api-reference/computers/stop",
   "quote": "Both modes preserve disk data; stopping discards live processes and memory, and starting cold-boots with potentially different connection details.",
   "note": "Stop (manual or per-computer auto_stop_minutes idle timer) archives the disk to object storage and cold-boots on a new host: persistent disk, not a state snapshot. A separate 'suspended' status exists ('Paused on its host. It resumes on interaction') but what triggers it and whether it survives host loss is not documented.",
   "checked": "2026-09-29"
  },
  "snapshot_on_demand": {
   "value": true,
   "source": "https://docs.orgo.ai/api-reference/computers/clone",
   "quote": "Cloning copies the full disk: installed software, files, browser sessions. The clone gets a new UUID, a new `instance_id`, and a fresh network identity.",
   "note": "POST /computers/{id}/clone of a running computer = disk-only fork into a new running computer (docs map Fly/DO/Hetzner 'snapshot' to clone). No named snapshot / restore-in-place endpoint in the API list; CLI has '--snapshot <id>' ('Boot from a snapshot or custom image') but no documented way to create one. Template builds produce memory+CPU 'golden snapshots' but from a spec, not from an arbitrary running computer.",
   "checked": "2026-09-29"
  },
  "own_agent_api": {
   "value": true,
   "source": "https://docs.orgo.ai/api-reference/chat/completions",
   "quote": "Send a message to an AI model and have it drive an Orgo computer on your behalf. The model sees the screen, clicks, types, and runs shell commands until your request is satisfied, then returns the final assistant message.",
   "note": "Orgo's own computer-use agent (Claude Sonnet/Opus models only) via OpenAI-compatible POST /v1/chat/completions with SSE streaming, plus /v1/threads list/get (history must be resent client-side); CLI 'orgo agent' / 'orgo run' / 'orgo resume'. The agent loop runs on Orgo's side and drives the computer, not inside it. Claude Code / Hermes / OpenClaw run inside via templates but there is no first-party API to prompt them.",
   "checked": "2026-09-29"
  },
  "harness_api": {
   "value": false,
   "source": "https://docs.orgo.ai/api-reference/chat/completions",
   "quote": "Send a message to an AI model and have it drive an Orgo computer on your behalf. The model sees the screen, clicks, types, and runs shell commands until your request is satisfied, then returns the final assistant message.",
   "note": "Orgo's own computer-use agent (Claude Sonnet/Opus models only) via OpenAI-compatible POST /v1/chat/completions with SSE streaming, plus /v1/threads list/get (history must be resent client-side); CLI 'orgo agent' / 'orgo run' / 'orgo resume'. The agent loop runs on Orgo's side and drives the computer, not inside it. Claude Code / Hermes / OpenClaw run inside via templates but there is no first-party API to prompt them.",
   "checked": "2026-09-29"
  },
  "egress_allowlist": {
   "value": true,
   "source": "https://docs.orgo.ai/guides/templates/schema",
   "quote": "Rules are domain patterns, exact domains, IPs, or CIDRs. Domain patterns match at the DNS layer, so `*.github.com` also covers subdomains. The VNC and Desktop API paths are always reachable. Omit the block entirely for no filtering.",
   "note": "Template field egress_policy, 'Per-computer network filtering, enforced on the host', mode allow (default-deny) or block. Only documented as a template field, not on plain POST /computers.",
   "checked": "2026-09-29",
   "level": "both"
  },
  "egress_http_rules": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/templates/schema",
   "quote": "Rules are domain patterns, exact domains, IPs, or CIDRs.",
   "note": "egress_policy rules are host/IP only (DNS-layer matching); no method/path/header rules documented. Separate BYO upstream HTTP CONNECT proxy applies to browser traffic only.",
   "checked": "2026-09-29"
  },
  "ingress_rules": {
   "value": true,
   "source": "https://docs.orgo.ai/guides/migrations/digitalocean",
   "quote": "No public per-computer IP. Reach it via the authenticated proxy: `www.orgo.ai/api/desktops/{instance_id}/proxy/<path>`",
   "note": "Access only through Orgo's API-key/token-authenticated proxy; template app ports are private unless marked public ('Expose publicly. Defaults to `false`.' - templates/schema). No IP allowlists documented.",
   "checked": "2026-09-29"
  },
  "guest_firewall": {
   "value": true,
   "source": null,
   "quote": null,
   "note": "Full VM (QEMU per prior research) with root desktop; not documented.",
   "checked": "2026-09-29",
   "derived": true
  },
  "secret_proxy": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/templates/secrets",
   "quote": "At create, Orgo writes the user's matching secrets into the computer at `/root/.env` (mode `0600`). They are never baked into the shared snapshot.",
   "note": "Vault secrets are injected as plain env/files inside the computer (also POST /computers/{id}/secrets/sync); the agent can read them. No credential-injecting proxy.",
   "checked": "2026-09-29"
  },
  "docker_inside": {
   "value": true,
   "source": "https://docs.orgo.ai/guides/instance-types",
   "quote": "Every Orgo computer runs on a dedicated instance with guaranteed resources.",
   "note": "derived: computers are full root Ubuntu VMs (own kernel, hypervisor field on the computer record) reachable over bash/SSH, and docs show apt-based installs of arbitrary software (Node, ffmpeg, systemd services); Docker itself is not named in any doc, so this is inferred from the VM architecture, not a direct statement.",
   "checked": "2026-09-30"
  },
  "anti_bot": {
   "value": false,
   "source": "https://www.orgo.ai/pricing",
   "quote": "Launch cloud computers that AI agents can control and interact with. Every Orgo computer is a full desktop environment with a browser and the standard userland.",
   "note": "not offered: no stealth/fingerprint-evasion feature documented anywhere; egress is plain data-center IPs (residential only in an allowlisted beta upstream proxy), with no anti-bot-detection product mentioned.",
   "checked": "2026-09-30"
  },
  "captcha_solving": {
   "value": false,
   "source": "https://www.orgo.ai/pricing",
   "quote": "Launch cloud computers that AI agents can control and interact with. Every Orgo computer is a full desktop environment with a browser and the standard userland.",
   "note": "not offered: Orgo is a general computer-use/agent-sandbox product (screenshot+click+type API); no CAPTCHA-solving feature, endpoint, or partner integration appears anywhere in the docs or changelog.",
   "checked": "2026-09-30"
  },
  "volume_attach": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/instance-types",
   "quote": "A computer's size is the `ram` and `cpu` you send to `POST /computers`; the names below are shorthand for common pairs, not API values.",
   "note": "not offered: a computer has one built-in persistent disk sized implicitly by the plan's storage pool; there is no separate volume-create/attach API or field anywhere in the API reference or templates schema.",
   "checked": "2026-09-30"
  },
  "volume_shared": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/instance-types",
   "quote": "A computer's size is the `ram` and `cpu` you send to `POST /computers`; the names below are shorthand for common pairs, not API values.",
   "note": "not offered: outside the product's scope — there is no volume/disk resource distinct from a computer's own disk, so nothing can be mounted by multiple running computers at once.",
   "checked": "2026-09-30"
  },
  "lang_python": {
   "value": true,
   "source": "https://docs.orgo.ai/api-reference/computers/exec",
   "quote": "For Python code execution, use Execute Python instead. POST /computers/{id}/exec",
   "note": "First-party API/SDK action `computer.exec(code)` / `orgo_execute_python` runs Python code directly inside the computer; Python 3.11+ is also installed by default per the setup guides.",
   "checked": "2026-09-30"
  },
  "lang_node": {
   "value": true,
   "source": "https://docs.orgo.ai/llms-full.txt",
   "quote": "Open orgo.ai/workspaces and create a computer with at least 8 GB RAM and 4 CPU cores. The installer pulls Python, Node.js, ripgrep, and ffmpeg.",
   "note": "Node.js is not exposed via a dedicated exec action (only Bash and Python have first-party actions), but it is installed by first-party setup scripts/templates (e.g. Node 20+/24 LTS for the CLI and OpenClaw) and is runnable via the Bash action; no evidence it ships by default on a bare computer without running an installer.",
   "checked": "2026-09-30"
  },
  "public_ipv4": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/migrations/digitalocean",
   "quote": "No public per-computer IP. Reach it via the authenticated proxy: `www.orgo.ai/api/desktops/{instance_id}/proxy/<path>`",
   "note": null,
   "checked": "2026-09-30"
  },
  "inbound_https": {
   "value": true,
   "source": "https://www.orgo.ai/pricing.md",
   "quote": "Serve an app running inside a computer under your own domain, with automatic TLS (HTTP/WebSocket ingress). Linux-only. The number of **active** custom domains is capped per plan and enforced against the workspace owner.",
   "note": "Public HTTPS ingress is via custom domains only: Starter 0, Growth 1, Scale 3, Enterprise unlimited. Without one there is no public hostname: the OpenClaw guide says \"A computer has no public inbound hostname, so run gateways in outbound mode ... Webhooks cannot reach the gateway.\" (https://docs.orgo.ai/guides/openclaw), and the authenticated proxy (www.orgo.ai/api/desktops/{id}/proxy/<path>) needs an Orgo API key. Not on Starter.",
   "checked": "2026-10-14"
  },
  "max_session_h": {
   "value": "none",
   "source": "https://docs.orgo.ai/llms-full.txt",
   "quote": "Linux computers boot in under 500 ms, and every computer runs continuously until you stop it.",
   "note": "auto_stop_minutes (0-1440) only controls an optional idle-timeout auto-stop; there is no maximum session/runtime cap documented — a computer runs indefinitely until the user stops it.",
   "checked": "2026-09-30"
  },
  "image_oci": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/templates/introduction",
   "quote": "Think of it as a `Dockerfile` for a full desktop computer: write it once, version it, and hand out reproducible computers from it.",
   "note": "Orgo is a full-VM desktop platform (isolation: vm); 'custom_image'/templates are Orgo's own YAML template format producing a golden VM snapshot, explicitly analogized to (not built on) Dockerfile/OCI. No Docker/OCI image import, registry, or 'docker pull'-style boot is documented anywhere in the API reference or CLI ('--snapshot <id>' boots Orgo's own snapshot/custom image format, not an OCI image).",
   "checked": "2026-09-30"
  },
  "static_egress_ip": {
   "value": true,
   "corrected_from": false,
   "source": "https://docs.orgo.ai/api-reference/computers/delete",
   "quote": "Deleting a computer does not release its dedicated egress IP, and does not stop the charge for it. ... Release it separately with `DELETE /api/egress-ips/{id}`, or reassign it to another computer with `PATCH /api/egress-ips/{id}`.",
   "note": "Orgo has a dedicated/static egress IP feature managed via /api/egress-ips (allocate, delete, and PATCH-reassign to a different computer), billed independently of the computer it's attached to. This is separate from the BYO browser-proxy feature already recorded (egress_ip_type note). Card's 'static_egress_ip: false' with no evidence was wrong.",
   "checked": "2026-09-30"
  },
  "arm64": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/templates/schema",
   "quote": "`cpu` | integer | `1`, `2`, `4`, `8`, `16` | vCPU count. Any other value is rejected. Capped by your plan.",
   "note": "No ARM/Graviton/Apple-Silicon CPU architecture option appears anywhere in the ~13,500-line combined docs (llms-full.txt), the instance-types/template schema (only a vCPU-count knob, no arch field), or the pricing page; GPU support (T4/L4/A10/L40S/A100/H100) further implies a CUDA/x86_64 host fleet. Absence is well-documented but not an explicit 'x86-only' statement, hence medium confidence.",
   "checked": "2026-09-30"
  },
  "byoc": {
   "value": true,
   "corrected_from": false,
   "source": "https://www.orgo.ai/pricing",
   "quote": "Enterprise Best for large fleets, custom needs Custom Volume pricing, billed annually. 50+ computers Up to 64 GB RAM per computer 50+ team members Linux + Windows + Mac Shared, dedicated, or on-prem Dedicated support with SLA",
   "note": "The Enterprise tier explicitly offers on-prem deployment (i.e. running Orgo's stack in the customer's own infrastructure/cloud), which is BYOC. Card's 'byoc: false' with no evidence was wrong; price is custom/sales-gated so exact terms aren't public, but the offering itself is documented.",
   "checked": "2026-10-14"
  },
  "open_source": {
   "value": false,
   "source": "https://www.orgo.ai/pricing",
   "quote": "Start with one computer for $29/month. Grow to a dedicated fleet on Enterprise.",
   "note": "Orgo is a closed commercial SaaS (paid monthly/yearly plans, sales-gated Enterprise, proprietary dashboard/API); only the thin client SDKs (`pip install orgo`, `npm install orgo`) are published as API wrappers, and even those aren't asserted to be open-source licensed in the docs. No OSS license, public core-platform repo, or self-hostable-for-free open-source release is documented for the Orgo platform itself, so 'right' at medium confidence (absence of any open-source claim, not an explicit 'closed source' statement).",
   "checked": "2026-09-30"
  },
  "eu_data_residency": {
   "value": false,
   "source": "https://docs.orgo.ai/guides/templates/schema",
   "quote": "`region` | string | `auto-us`, `iad`, `sjc`, `lax` | Accepted, but not used for placement.",
   "note": "Only US region codes exist (template field, and the CLI `--region` is a \"Region hint\"); no EU code or data-residency statement anywhere.",
   "checked": "2026-10-14"
  },
  "usage_api": {
   "value": false,
   "source": "https://docs.orgo.ai/llms-full.txt",
   "quote": "Your Usage page (profile dropdown -> Usage) shows current consumption and lets you set custom limits. Orgo bills per plan (a bundle of computers, a RAM pool, and support). Computers stay online until you stop them; no per-hour charge published.",
   "note": "granularity: account-only; cost: no; Flat monthly billing by plan tier, not per-sandbox. The dashboard Usage page exists but has no public API. The /usage endpoint only returns AI model token usage, not compute/VM billable metrics.",
   "confidence": "high",
   "checked": "2026-10-01"
  },
  "custom_domain": {
   "value": true,
   "source": "https://www.orgo.ai/pricing.md",
   "quote": "Serve an app running inside a computer under your own domain, with automatic TLS (HTTP/WebSocket ingress). Linux-only.",
   "note": "Active custom domains per plan: Starter 0, Growth 1, Scale 3, Enterprise unlimited. Only stated on the pricing price list; no docs page describes setup.",
   "checked": "2026-10-14"
  },
  "windows": {
   "value": true,
   "source": "https://docs.orgo.ai/api-reference/computers/create",
   "quote": "`windows` needs a Scale plan or a purchased Windows licence, otherwise it returns `403` with code `WINDOWS_REQUIRES_SCALE`.",
   "note": "Pricing page: \"Linux on every plan, Windows on Scale or as an add-on, and Mac and iPhone by request.\" Licence $15/mo per Windows computer on Starter/Growth (pricing.md).",
   "checked": "2026-10-14"
  },
  "gpu": {
   "value": true,
   "source": "https://docs.orgo.ai/api-reference/computers/create",
   "quote": "vGPU slice to attach: `2q` (2 GB VRAM) or `4q` (4 GB VRAM).",
   "note": "Self-serve on Linux only (\"combining it with another `os` returns `400` with code `GPU_REQUIRES_LINUX`\"); if no GPU host is free the create returns 503 or silently falls back to CPU (gpu_downgraded). GPU model and price unpublished. Template `gpu` (t4..h100) is \"Accepted, but not used at launch.\" (templates/schema).",
   "checked": "2026-10-14"
  },
  "android": {
   "value": true,
   "source": "https://docs.orgo.ai/api-reference/computers/create",
   "quote": "Operating system: `linux`, `windows`, `macos`, `android`, or `ios`. Any other value returns `400`.",
   "note": "\"Every value other than `linux` only succeeds where the fleet has a host that supports it, otherwise create returns `422` with code `OS_UNAVAILABLE`.\" `ios` is a physical handset ($65/mo by request), not a VM.",
   "checked": "2026-10-14"
  },
  "macos": {
   "value": true,
   "source": "https://www.orgo.ai/pricing",
   "quote": "Linux on every plan, Windows on Scale or as an add-on, and Mac and iPhone by request.",
   "note": "By request only (Enterprise or an approved Custom machines request); Mac from $90/mo per pricing.md. macOS RAM fixed at 10 GB (MACOS_RAM_FIXED).",
   "checked": "2026-10-14"
  }
 },
 "locations": [
  {
   "country": "US",
   "state": "VA",
   "city": "Ashburn",
   "provider_region": "iad",
   "self_serve": true,
   "choosable": false
  },
  {
   "country": "US",
   "state": "CA",
   "city": "San Jose",
   "provider_region": "sjc",
   "self_serve": true,
   "choosable": false
  },
  {
   "country": "US",
   "state": "CA",
   "city": "Los Angeles",
   "provider_region": "lax",
   "self_serve": true,
   "choosable": false
  }
 ],
 "locations_source": "https://docs.orgo.ai/guides/templates/schema",
 "locations_quote": "`region` | string | `auto-us`, `iad`, `sjc`, `lax` | Accepted, but not used for placement.",
 "locations_checked": "2026-10-14",
 "locations_note": "Cities come from the documented US region codes (iad, sjc, lax). The template field is ignored for placement and the CLI flag is only a \"Region hint\", so a location cannot be chosen; Orgo does not state where computers actually run beyond US."
}