# Dedalus Labs Official pricing: https://www.dedaluslabs.ai/pricing Category: agent-sandbox · Isolation: vm ## Pricing regimes (raw) - **On-demand** (resource): $0.04536/vCPU-h, $0.01458/GiB-h [beta] ## Features Yes: pause/resume, persistent disk, idle auto-stop, ≥24 h sessions, full VM (own kernel), Docker inside, root, systemd, SSH, BYOC, SSO, Python, wake on request, live resize, audit logs, gVisor or VM (no shared kernel), firewall inside (nftables) No: snapshots, memory snapshots, fork/clone, volumes, custom image (Docker or snapshot), start from your own snapshot, your own Docker/OCI image, browser, desktop GUI, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, residential IPs, preinstalled agents, public IPv4, HTTPS preview URLs, egress allowlist, static egress IP, self-hosting, open source, SOC 2, GPU, arm64, Windows, macOS, HTTP method/path egress rules, live fork (no pause), memory fork, EU data residency, MCP server, automatic snapshots, snapshots on demand, agent harness API, hosted agent API (their own agent), inbound access rules, secret proxy, secret proxy for any API, extra volumes, shared volumes Unknown: everything else. Evidence (source + quote) per feature: https://battleships.dev/data/providers/dedalus-labs.json → feature_evidence ## Caveats - Cloud Hypervisor-derived custom VMM; CPU/RAM charged per second while actively running, idle sleep free. This is lifecycle-active, not proven CPU-utilization billing. - Storage headline incorrectly labels $0.073 as per GiB-hour; calculator/FAQ explicitly use $0.0001/GiB-hour ($0.073 at 730h). Use FAQ and preserve conflict. Unlimited Pro storage means capacity, not free bytes. - Hobby 5 machines and 50 compute-hours/month; Pro 20 machines. Fifty concurrent requires Enterprise, whose fee and rates are unknown. Pricing footer says join waitlist: public rates do not prove instant admission. - Null values mean unknown, not free/unlimited. No model-token cost, tax or unverified ancillary charge included. - Native browser/device/task/credit prices with unknown CPU/RAM are deliberately non-priceable. Check regime before interpreting any partial resource estimate. - SSH verified 2026-09-28: skill docs say "Use dedalus ssh $MACHINE_ID for an interactive shell" (https://docs.dedaluslabs.ai/.well-known/skills/dedalus-machines/skill.md). - Lifecycle mutations are rate limited (RATE_LIMITED, scope mutating_lifecycle_routes, retry_after_ms) with no published numeric limit. - SSH sessions close after 30 idle minutes or 12 hours total; deleted machines' storage may be retained up to 30 days. - Autosleep default is 300 idle seconds ('Machines sleep after five minutes of inactivity by default'); pricing says 'Idle time is free' and billing is for active compute, but docs say compute billing stops on sleep. Whether the up-to-5-minute idle tail before autosleep is billed is not stated. ## How this provider charges YC: Summer 2025; directory status **Active**. Research scope: public-product-researched. Source: https://www.ycombinator.com/companies/dedalus-labs Compute substrate for AI agents ## Regime table | Regime | When it applies | How billed | Numbers | Source | |---|---|---|---|---| | On-demand | Public listed mode | resource; CPU alloc, memory alloc | $0.04536/vCPU-h + $0.01458/GiB-h; multiplier 1 | https://www.dedaluslabs.ai/pricing | | Hobby | Plan limits apply | Monthly fee; credits only as stated | fee=0; included usage=$0; concurrent=5; max session h=None | https://www.dedaluslabs.ai/pricing | | Pro | Plan limits apply | Monthly fee; credits only as stated | fee=20; included usage=$20; concurrent=20; max session h=None | https://www.dedaluslabs.ai/pricing | | Enterprise | Plan limits apply | Monthly fee; credits only as stated | fee=None; included usage=$None; concurrent=None; max session h=None | https://www.dedaluslabs.ai/pricing | | Retained disk / GiB-month | Lifecycle/scope must be checked | Native add-on | $0.073 | https://www.dedaluslabs.ai/pricing | | Snapshots / GiB-month | Lifecycle/scope must be checked | Native add-on | unverified, not $0 | https://www.dedaluslabs.ai/pricing | | Egress / GiB | Lifecycle/scope must be checked | Native add-on | unverified, not $0 | https://www.dedaluslabs.ai/pricing | | IPv4 / month | Lifecycle/scope must be checked | Native add-on | unverified, not $0 | https://www.dedaluslabs.ai/pricing | ## Verified billing facts and gotchas 1. Cloud Hypervisor-derived custom VMM; CPU/RAM charged per second while actively running, idle sleep free. This is lifecycle-active, not proven CPU-utilization billing. 2. Storage headline incorrectly labels $0.073 as per GiB-hour; calculator/FAQ explicitly use $0.0001/GiB-hour ($0.073 at 730h). Use FAQ and preserve conflict. Unlimited Pro storage means capacity, not free bytes. 3. Hobby 5 machines and 50 compute-hours/month; Pro 20 machines. Fifty concurrent requires Enterprise, whose fee and rates are unknown. Pricing footer says join waitlist: public rates do not prove instant admission. ## Worked example Requested: 4 vCPU/8 GiB, 50 simultaneous × 8 h/day × 22 days = **8,800 instance-hours**, 30% CPU; 50 GiB snapshots and 100 GiB egress. A month is 730 hours only when explicitly used in rate conversion. List-rate compute illustration: 8,800×(4×0.04536+8×0.01458)=$2,623.10. 50 GiB storage at $0.0001/GiB-h ×730 adds $3.65. Not a purchasable total: Pro supports only 20 machines; 50 concurrent needs an Enterprise quote. Do not assume 30% CPU utilization yields 70% savings. ## Sources and coverage - https://www.ycombinator.com/companies/dedalus-labs - https://www.dedaluslabs.ai/pricing - https://www.dedaluslabs.ai/blog/dedalus-machines-journey - https://docs.dedaluslabs.ai/.well-known/skills/dedalus-machines/skill.md - https://www.dedaluslabs.ai/ - https://www.dedaluslabs.ai/llms.txt Page captures and documentation indexes/bundles are in `raw/`; provider JSON lists the evidence paths. HN query results were captured separately as `--hn.txt`; they are discovery leads, not current tariff authority. Downloaded docs do not imply every feature was verified; unsupported assertions remain null.