FAQ

🌐 General Questions

What is Meshive?

Meshive is a task-first GPU cloud for AI training, inference, rendering, and serverless model serving. You get Docker containers with exclusive access to NVIDIA GPUs, and can scale from a single RTX 4090 to a cluster of H100s instantly.

How does the marketplace work?

Meshive acts as a bridge between high-tier data centers and developers. You can browse available hardware based on performance, location, and price, then deploy a pre-configured environment in seconds.

Is my data secure on Meshive?

Yes. Each instance runs in an isolated Docker container. While we recommend regular backups for critical data, our infrastructure is designed to ensure that your workspace remains private and your processes are shielded from other users on the network.

---

💳 Billing & Payments

How does the billing system work?

We believe in fairness. Meshive bills in fixed one-minute intervals — fees are calculated every 60 seconds based on the pods you have running, and credits are deducted from your account balance as billing rolls. No minimums, no monthly commitments.

Am I billed while my host is under maintenance?

No. While a host’s maintenance window is open, your pods and volumes on that machine aren’t billed. Meshive doesn’t stop your pods — they keep running unless the host’s work interrupts them. Billing resumes when the window ends (the host may end it early). Pods on other machines that mount a network volume on a machine in maintenance are still billed. Tasks and Assets storage are unaffected. See Pricing & Instances.

Is there free credit?

There is no sign-up credit. GPU pods, tasks and Serverless inference all draw from the same account balance, so add credit before you start. You can earn credit by inviting friends: when someone you referred makes their first top-up, you receive a share of it as credit (see Account → Referral in the console). Referral credit can be used for GPU pods, tasks and Serverless inference, and doesn’t expire.

What payment methods do you accept?

We currently support major credit cards via Stripe. We are also working on integrating cryptocurrency payments to better serve our global community.

Are there any hidden fees?

No. The price you see on the marketplace includes GPU, CPU, and RAM. Note that minimal charges apply for storage persistence even when the instance is stopped, as the data remains on the physical disk. The exception is host maintenance: a volume on a machine with an open maintenance window isn’t billed until the window ends.

---

🚀 Instances & Connectivity

What GPUs are available?

Our fleet includes a wide range of NVIDIA hardware, including:

• Consumer Grade: RTX 4090, 3090.
• Enterprise Grade: H100, A100 (80GB), A6000, and L40S.

How do I connect to my instance?

Meshive supports multiple entry points:

• SSH — direct terminal access for power users.
• Jupyter Notebook — pre-configured for data science and rapid prototyping.
• VS Code (Remote) — connect your local IDE to the Meshive remote environment via SSH.

Can I use my own Docker image?

Absolutely. You can point to any public image on Docker Hub or use your private registry. Meshive also provides optimized “One-Click” templates for PyTorch, TensorFlow, and Stable Diffusion.

---

🔧 Troubleshooting

My instance is ‘Initializing’ for a long time. What should I do?

This usually happens during the initial pull of a large Docker image. If it takes longer than 10 minutes, check your internet connection or try a different host. You can also view the System Logs in the Meshive Console for specific error messages.

Can I upgrade my GPU without losing data?

Currently, instances are tied to specific hardware. To move to a more powerful GPU, we recommend saving your progress to a Docker image or using a cloud storage mount (like S3) to transfer your weights and datasets.

How do I get technical support?

If you encounter any issues, you can:

1. Join our Discord Community for real-time help from other developers.
2. Contact us at [email protected].