Y Combinator lists TensorPool as acquired. Current hosted onboarding, resource prices and retention terms are not verified; no universal service shutdown date is established by that status. Read the source ↗

GPU clouds / TensorPool GPU Orchestration alternative

A TensorPool alternative for the agent around GPU clusters

TensorPool’s public CLI describes job submission and on-demand multi-node GPU clusters with storage-management commands. Y Combinator currently lists the company as acquired. That status does not by itself prove every hosted service has ended, but current account access and commercial terms need confirmation before treating it as a new-service option.

Start buildingCompare capabilities ↓

Reviewed 2026-10-19 · Published by the Sandbox0 team

When TensorPool GPU Orchestration fits

TensorPool’s surviving CLI and launch materials can help understand an existing GPU-job workflow. Confirm hosted cluster access, retained volumes and the effective agreement with the provider before relying on continued service.

When to consider Sandbox0

Consider Sandbox0 for compatible Linux agent preparation and orchestration, with durable tools and configuration across the whole writable RootFS, automatic online growth and supported files while compute stays paused. Required GPU clusters remain separate.

Capabilities

TensorPool GPU Orchestration and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeTensorPool GPU OrchestrationNot established in this reviewSandbox0ttl and hard_ttl default to 0 (disabled). Configure idle pause and deletion policies for the same persistent sandbox identity; an idle running sandbox still accrues compute charges.Read documentation →
Persistent filesTensorPool GPU OrchestrationPublic storage/NFS command groups exist, but current hosted volume retention and access are not established by the surviving CLI alone.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSTensorPool GPU OrchestrationJob and cluster orchestration do not establish current full writable OS filesystem continuity.View evidence ↗Sandbox0Persistence extends beyond /workspace to paths such as /usr/local, /etc, and /home. Runtime tmpfs and virtual filesystems are excluded.Read documentation →
Release compute, keep stateTensorPool GPU OrchestrationThe current public company profile and CLI do not establish a durable resumable hosted agent-machine lifecycle.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateTensorPool GPU OrchestrationThe reviewed public interface does not establish arbitrary RAM/process checkpoint restoration.View evidence ↗Sandbox0Experimental memory checkpoints can preserve compatible process state. Use memory: true on both pause and resume on a compatible deployment. A filesystem checkpoint alone does not preserve running processes.Read documentation →

Online RootFS growth

On growth-enabled images, Sandbox0 automatically expands the writable RootFS as free space runs low, within a configured sparse logical ceiling up to 64 TiB. Agent processes keep running. Growth is enabled in the qualified Cloud region; confirm activation for other regions. Existing images and snapshots retain their captured storage geometry. The ceiling is logical capacity, not preallocated host disk.

The surviving cluster and storage CLI does not establish automatic online growth of a retained per-agent RootFS. Confirm existing cluster storage behavior separately from the agent execution environment.

Files while compute is paused

Sandbox0's Files API reads persistent files without resuming compute. Filesystem-only pauses also permit edits; retained-memory pauses are read-only and reject mutations with 409. SSH, SFTP, file watching, and runtime-only mounts require running compute.

See paused file operations →

Acquisition and hosted availability are different facts

The YC company profile marks TensorPool Acquired and describes its CLI-oriented training workflow. It does not provide a universal shutdown date or a replacement service agreement. Check the actual account and storage access rather than concluding that acquisition alone means data is gone.

Source 1 ↗

The public CLI separates jobs, clusters and storage

The repository exposes cluster, storage, NFS, job and SSH command groups. Those interfaces establish the intended GPU workflow, but they do not establish a current whole mutable OS persistence or saved-RAM contract. Export needed job configurations, data locations and result files before changing execution providers.

Source 1 ↗

Current new-service pricing is not established

A current numeric cluster/storage tariff and hosted onboarding terms remain unverified. Historical GPU and storage rates may no longer apply. Request the applicable cluster, retained-data and support agreement if an existing account remains active.

Source 1 ↗Source 2 ↗

Pricing

Compare billing models, then price your workload

CPU allocation, memory, billing minimums, disk, plan fees, and supported workload differ. These meters are not equal-performance machine quotes. Confirm region, resource units, and included allowances before estimating a total.

TensorPool GPU Orchestration

Pricing reviewed 2026-10-19

Current GPU cluster tariff
Confirm the effective provider agreement
Rate source ↗
Current retained storage tariff
Confirm current retained-storage pricing and access
Rate source ↗
Existing account continuity
Provider confirmation required
Rate source ↗

No archived GPU floor or early-user launch credit is presented as a current offer. Confirm acquisition-related service arrangements, storage access and complete resource terms; acquisition is not itself a verified shutdown notice.

Check current TensorPool GPU Orchestration pricing ↗

Sandbox0

$0.015/GiB-hour compute (billed by the second at approximately $0.000004167/GiB-second), plus $0.02/GiB-month of retained incremental RootFS and memory checkpoint data. Shared base images and platform recovery retention are excluded; paused sandboxes accrue storage charges only.

Compute / GiB-hour
$0.015
Retained data / GiB-month
$0.02
Paused compute
$0.00

A 4 GiB agent, 100 running hours

$6.00 compute

4 GiB × 100 hours, billed by the second. Keeping 20 GiB of billable incremental data for a full billing month adds $0.40. CPU is derived from memory by the platform; choose a tier that fits your workload.

Understand storage billing →

Before migrating from TensorPool GPU Orchestration

  1. Confirm existing account access, acquisition-related service arrangements and whether clusters, volumes and job results remain accessible.
  2. Export verified code, job definitions, data and results; retain or replace suitable GPU capacity separately.
  3. Move compatible Linux agent stages and test full-system retention, RootFS growth, supported paused files and the new complete resource budget.

Common questions

Is Sandbox0 a drop-in TensorPool GPU Orchestration replacement?

Evaluate API compatibility and product scope. These products serve different core workflows, so a change may require a different architecture.

Does keeping files also keep a running agent?

Filesystem state and volatile process state are separate. Sandbox0 requires explicit compatible memory checkpoints to preserve running processes; its default filesystem resume starts fresh processes.

Sources and review scope

The Boat team's Battleships directory is the discovery source. References below describe the product and plan discussed on this page. Check current availability, regional terms and billing units before migrating. Where the cited material does not establish a capability or price, the comparison identifies that uncertainty.

Battleships research record ↗www.ycombinator.com/companies/tensorpool ↗github.com/tensorpool/tensorpool ↗tensorpool.dev/pricing ↗Sandbox0 RootFS capacity →

Related comparisons

Azure GPU VMs (NC_A100_v4 reference) vs Sandbox0 →Cerebrium Serverless Compute vs Sandbox0 →Crusoe vs Sandbox0 →DigitalOcean GPU Droplets vs Sandbox0 →Fluidstack vs Sandbox0 →Google Compute Engine GPU VMs vs Sandbox0 →