Blaxel
Blaxel is a managed cloud platform that provisions one microVM per AI agent, app, or job so that LLM-generated code can execute in isolation. Each sandbox auto-suspends when idle and resumes in ~25 ms with both filesystem and in-memory process state preserved. It is used to host coding-agent harnesses such as the Claude Agent SDK, OpenAI Codex CLI, Cline, and Cursor background agents.
| Pricing model | Per-second active compute billed by RAM-seconds (storage and snapshots billed separately per GB-month) |
|---|---|
| Pricing The headline compute price as the provider states it. Sorting ranks prices stated per vCPU over time, converted to hourly. Other pricing models sort after, unranked. | $0.0000115 per GB-RAM-second (active CPU) |
| Free while idle Whether a stopped, paused, or sleeping sandbox costs nothing. | ✓ |
| Elastic Resources flex with what the sandbox actually uses, and the bill follows: yes when the provider bills on active or observed use, no when capacity is reserved or allocated and billed while it runs. | ✓ |
| Memory snapshots Whether RAM state survives a pause and resume, not just disk. | ✓ |
| Wake on request Whether a stopped sandbox wakes automatically on inbound traffic. | ✓ |
| Start / resume Typical time from create or resume to running. Sorting ranks stated times, converted to seconds; claims with no figure sort after, unranked. | ~25 ms (resume from standby) |
| Max runtime The longest a sandbox may run. | Tier-dependent: up to 7 days (tier 0), up to 30 days (tier 1); unlimited in tier 2 and above |
| Isolation The isolation technology between sandboxes. | microVM with hardware-level (kernel-level) isolation |
| GPUs Whether GPU instances are available for sandboxes. | ✕ |
| Docker Whether Docker containers can run inside the sandbox. | ✓ |
| SDKs | Python, TypeScript, Go |
✓ yes · ✕ no · – no cited public fact.
Blaxel offers $200 in free credits on signup and supports Zero Data Retention (ZDR) natively. The platform is CPU-only, with its own blog stating it is 'not suitable for GPU workloads.' Pricing is RAM-based rather than vCPU-based, which is unusual relative to most sandbox providers.
Sources
- https://blaxel.ai/ai-information
- https://docs.blaxel.ai/sdk-reference/introduction
- https://blaxel.ai/sandbox
- https://blaxel.ai/
- https://blaxel.ai/blog/best-platforms-high-concurrency-sandbox-environments
- https://blaxel.ai/blog/soc-2-compliant-sandbox-providers
- https://blaxel.ai/pricing
- https://docs.blaxel.ai/Sandboxes/best-practices
- https://docs.blaxel.ai/Sandboxes/Expiration
- https://docs.blaxel.ai/Sandboxes/Overview
- https://blaxel.ai/blog/vercel-sandbox-alternatives
- https://docs.blaxel.ai/Overview
- https://blaxel.ai/blog/serverless-gpu-platforms-2026
- https://blaxel.ai/blog/how-to-fix-cold-start-latency-for-ai-agents
Last verified 2026-08-03. Corrections land through the update log.