# OpenPond Cloud Sandboxes

Official pricing: https://openpond.ai/pricing  
Category: agent-sandbox · Isolation: firecracker

## Pricing regimes (raw)

- **On-demand** (resource): $0.072/vCPU-h, $0.0234/GiB-h: Tiny (0.25 vCPU / 0.5 GiB / 4 GiB) 0.25 vCPU/0.5 GiB $0.03018/h; Small (0.5 vCPU / 1 GiB / 8 GiB) 0.5 vCPU/1 GiB $0.06036/h; Default (1 vCPU / 2 GiB / 10 GiB) 1 vCPU/2 GiB $0.12/h; Builder (2 vCPU / 4 GiB / 20 GiB) 2 vCPU/4 GiB $0.24/h; Heavy Builder (4 vCPU / 8 GiB / 40 GiB) 4 vCPU/8 GiB $0.48/h

## Features

Yes: snapshots, fork/clone, pause/resume, persistent disk, volumes, idle auto-stop, ≥24 h sessions, custom image (Docker or snapshot), start from your own snapshot, your own Docker/OCI image, full VM (own kernel), Docker inside, HTTPS preview URLs, egress allowlist, open source, SSO, Node.js, live fork (no pause), scheduled runs (cron), scoped API keys, spend limits, MCP server, snapshots on demand, hosted agent API (their own agent), gVisor or VM (no shared kernel), inbound access rules, firewall inside (nftables), extra volumes, FUSE mounts, kernel features (eBPF, mounts, cgroups)

No: memory snapshots, nested virtualization, browser, desktop GUI, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, residential IPs, preinstalled agents, SSH, public IPv4, static egress IP, self-hosting, BYOC, GPU, arm64, Windows, macOS, Python, HTTP method/path egress rules, wake on request, live resize, memory fork, usage per sandbox (API), automatic snapshots, agent harness API, secret proxy, secret proxy for any API, shared volumes

Unknown: everything else. Evidence (source + quote) per feature: https://battleships.dev/data/providers/openpond.json → feature_evidence

## Company

- Ownership: independent
- Stage: unknown
- Accelerators: none found
- Total raised: unknown
- Revenue: unknown

OpenPond appears to be a small, early-stage, founder-led company (Lost Labs LLC) with two named co-founders (Adam Elhammamy, Sam Cesario) and no publicly disclosed institutional funding, accelerator participation, investors, or revenue figures as of 2026-09-30. Product is an open-source AI-agent 'Harness' plus a hosted cloud sandbox / managed RL offering. A Crunchbase profile URL for 'openpond' returned 403 (blocked, not confirmed to exist) and no Y Combinator company page was found at the expected slug. Buyers should treat OpenPond as financially unverified/early-stage rather than assume any specific funding stage.

## Caveats

- Priced from the published per-resource meters ($0.072/vCPU-h, $0.0234/GiB-h memory, $0.00012/GiB-h VM disk; https://openpond.ai/docs/sandbox-sdk/pricing-and-budgets, 2026-10-14). VM disk is billed per GiB only while running ("Stopping ends active compute billing but intentionally retains recoverable sandbox state."), so it is modelled as $0.0876/GiB-month of running time with no included disk (every GiB counts, the 10 GiB default adds $0.0012/h). Durable storage $0.0518/GiB-month (used for retained state). Monthly usage credits ($10 Starter / $20 Pro) also cover model calls. Egress unpublished.
- Features set to null were not verified. No sandbox was purchased or tested; feature/performance statements are documentation claims. Null maximums are unknown unless a cited note explicitly states unlimited.
- A complete monthly bill requires known snapshot/egress/plan limits; do not silently treat missing ancillary charges as free.
- Rate card is served with pricingSource "openpond_poc_config" in the SDK types, and docs say 'The authenticated Billing page and the SDK pricing endpoint are authoritative for the rate card applied to your account'. Treat the published retail rates as possibly provisional.
- Size ceiling 8 vCPU / 32 GiB RAM / 100 GiB disk comes from the open-source template manifest schema ("diskGb: z.number().positive().max(100).optional()"); the hosted control plane may apply different limits to SDK-created sandboxes. The engine has no per-sandbox disk maximum, so disk requests above 100 GiB are not flagged.
- Whether retained state of a stopped sandbox (its VM disk) is billed as storage after stop is not stated: the docs say only that "Snapshots, archives, and durable volumes can continue consuming retained-storage quota or storage charges after compute stops." The card treats stopped VM disk as unbilled.

## How this provider charges


As of 2026-09-28. Prices are USD unless explicitly labelled otherwise. Source type: official published list/promo or unknown, never a fabricated quote.

Hosted sandbox plans with monthly shared usage credits. CPU/RAM unit rates not found. Team subscription is not an unlimited compute pool.

| Regime | When it applies | How billed / numbers | Source |
|---|---|---|---|
| Free | 1 concurrent | Limited included monthly usage; top-ups available. | https://openpond.ai/pricing |
| Starter | $10/month | $10 usage;10 concurrent;3 members included, extra$8/member. | https://openpond.ai/pricing |
| Pro | $20/month | $20 usage;10 concurrent;3 members included, extra$8/member. | https://openpond.ai/pricing |
| Enterprise | C50 workload | Custom dedicated capacity and usage policy. | https://openpond.ai/pricing |

## Gotchas

- Hosted sandbox plans with monthly shared usage credits. CPU/RAM unit rates not found. Team subscription is not an unlimited compute pool.
- Not fully priceable in ENGINE_CARD: unknown USD resource rates or machine shape. Null is not zero; do not rank as a free provider.
- Features set to null were not verified. No sandbox was purchased or tested; feature/performance statements are documentation claims. Null maximums are unknown unless a cited note explicitly states unlimited.
- A complete monthly bill requires known snapshot/egress/plan limits; do not silently treat missing ancillary charges as free.

## Worked workload

4 vCPU / 8 GiB; 50 simultaneous instances × 8 h/day × 22 days = **8,800 instance-hours**, **35,200 vCPU-hours**, **70,400 GiB-hours**. At 30% observed CPU:10,560 active-vCPU-hours only where the meter actually uses utilization. Assume one start/instance/day:1,100starts. Retain50GiB snapshots for a month;100GiB outbound. Active disk capacity was not specified.

Self-serve concurrency10 cannot serve50. Need Enterprise quote and resource-hour rates; cannot divide plan fees by8,800h to invent a rate.

## Sources

- https://openpond.ai/pricing
- https://github.com/openpond/openpond