Home

ai

Modal

Serverless compute platform for GPU inference, model training, batch processing, and containerized Python workloads. It bills resource consumption by execution time and provides a monthly included compute allowance on its Starter plan.

Data status: partially_verified · Checked: 2026-10-16

Good fit if

On-demand CPU, memory, GPU, volume, and network resources with scale-to-zero behavior for supported workloads.

May not fit if

Compute, GPU, memory, storage, egress, region multipliers, and non-preemptible execution have separate rates; shared endpoint tokens are not covered by Starter compute credits.

Price

0 USD

Pricing source ↗
Official website ↗

Features and benefits

  • On-demand CPU, memory, GPU, volume, and network resources with scale-to-zero behavior for supported workloads.
  • Starter plan has no monthly platform fee and includes $30 monthly compute credits, subject to eligible compute categories.

Limitations

  • Compute, GPU, memory, storage, egress, region multipliers, and non-preemptible execution have separate rates; shared endpoint tokens are not covered by Starter compute credits.
  • The Team plan adds a $250 monthly platform charge plus compute; actual totals depend on requested resources and runtime.

Pricing

Starter compute credits

0 USD · free

Starter has $0 platform fee and includes $30/month compute credits. Compute above the allowance and excluded token-priced shared endpoints can incur charges; this is not unlimited free use.

Pricing source

Nvidia T4 GPU compute

0.000164 USD · usage_based

Published rate per GPU-second; CPU, memory, storage, egress, region multipliers, and other GPU types are billed separately. This specific GPU rate is not a typical workload estimate.

Pricing source

Team platform fee

250 USD · monthly

Monthly platform fee plus compute charges; includes $100/month compute allowance under the terms listed on the pricing page.

Pricing source

Sources