Modal

Depot darling for Python agent backends and bursty GPU jobs. If you think in functions not VMs, start here.

Listing status · Claim or correct this listing

Depot darling for Python agent backends and bursty GPU jobs. If you think in functions not VMs, start here. Strengths: Best-in-class Python DX for serverless compute; GPUs + sandboxes + cron in one model; Scale-to-zero economics.

Overall

8.4

curated_estimate

confidence: high · as of 2026-10-19

Capability fit

9.0

Price / performance

8.2

Latency / reliability

8.6

Developer & agent DX

9.6

Privacy / data retention

7.8

Lock-in risk (higher = worse)

4.5

Maturity / track record

8.2

Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.

Sources: Modal: https://modal.com · Modal pricing: https://modal.com/pricing

Depot darling for Python agent backends and bursty GPU jobs. If you think in functions not VMs, start here.

Strengths

· Best-in-class Python DX for serverless compute
· GPUs + sandboxes + cron in one model
· Scale-to-zero economics

Weaknesses

· Premium vs raw RunPod $/GPU-sec sometimes
· Python-centric (polyglot harder)
· Less classic VM control

Pricing notes

As of 2026-10-24 (modal.com/pricing, fetched 2026-10-24): Python-native serverless GPU/CPU billed per second of compute ('You never pay for idle resources'). GPU list rates per second: H100 SXM5 '$0.001097', H200 SXM '$0.001261', B200 '$0.001736', A100 80 GB '$0.000694', L40S '$0.000542', A10 '$0.000306', L4 '$0.000222', T4 '$0.000164'; CPU '$0.0000131' per core per second (physical core, 2 vCPU equivalent); memory '$0.00000222' per GiB per second. Region selection '1.15 - 1.75x base prices'; non-preemptible execution '3x base prices' ('Sandboxes are billed at non-preemptible rates of 3x list price.'). Containers are kept alive '60 seconds after the last inputs', which is billed (configurable). Plans: Starter '$0 + compute / month' with '$30 / month free credits' ('100 containers + 10 GPU concurrency'); Team '$250 + compute / month' with '$100 / month free credits' ('5000 containers + 50 GPU concurrency'); Enterprise custom. Shared Endpoints are 'billed per token rather than per second' and not covered by included compute ('This took effect September 1, 2026.'). Excellent DX for jobs and endpoints; cold starts matter for interactive agents.

Regions: us, globalLast reviewed: 2026-10-19Status: active

Sourced pricing

Contract status unit_mismatch · unit_mismatch. Single normalized $/1M: null. Machine list: /api/pricing?provider=modal

Nvidia B300 · usd_per_gpu_second · $0.001972 · unit_mismatch · normalized null
Nvidia B200 · usd_per_gpu_second · $0.001736 · unit_mismatch · normalized null
Nvidia H200 SXM · usd_per_gpu_second · $0.001261 · unit_mismatch · normalized null
Nvidia H100 SXM5 · usd_per_gpu_second · $0.001097 · unit_mismatch · normalized null
Nvidia RTX PRO 6000 · usd_per_gpu_second · $0.000842 · unit_mismatch · normalized null
Nvidia A100 80GB · usd_per_gpu_second · $0.000694 · unit_mismatch · normalized null
Nvidia A100 40GB · usd_per_gpu_second · $0.000583 · unit_mismatch · normalized null
Nvidia L40S · usd_per_gpu_second · $0.000542 · unit_mismatch · normalized null
Nvidia A10 · usd_per_gpu_second · $0.000306 · unit_mismatch · normalized null
Nvidia L4 · usd_per_gpu_second · $0.000222 · unit_mismatch · normalized null
Nvidia T4 · usd_per_gpu_second · $0.000164 · unit_mismatch · normalized null
CPU physical core · usd_per_core_second · $0.0000131 · unit_mismatch · normalized null

Community signals

Insight claims only — never mutate official scores. Outside reports: 0; seed examples: 0.

0 insights

No insights yet. Submit a production claim →

All insights for this provider →