Beam Serverless GPU

Infrastructure service

Serverless GPU endpoints that turn a Python function into a model-serving endpoint on RTX 4090 and RTX 5090 GPUs that start when a request arrives and stop after traffic stops, billed by the millisecond.

Find alternatives to Beam Serverless GPU

TypeInfrastructure service
RoleBundled component
AvailabilitySold
DeploymentSaaS
Intended for—
SecurityNot recorded
StatusActive
Sold withinBeam — Included with the parent
PricingSubscription, Usage-based — official pricing ↗

What it does

beam.cloud

Sources

Pricing

Pricing page · 16 Nov 2026

Pricing page text fetched 2026-11-16 with curl; page quote: "RTX 4090 PCIE 24 GB VRAM".

Description

Company disclosure · 16 Nov 2026

Product page text fetched 2026-11-16 with curl; page quote: "Run your model on RTX 4090 and RTX 5090 GPUs that start when a request arrives and shut down after traffic stops.". The page says "GPU time starts at $0.69 /hr, metered by the millisecond"; the separate /inference page ("Serverless inference for any model") describes the same endpoint product, so it is not a second row.

What it does

Company disclosure · 16 Nov 2026

Product page text fetched 2026-11-16 with curl; page quote: "Run your model on RTX 4090 and RTX 5090 GPUs that start when a request arrives and shut down after traffic stops.". The page says "GPU time starts at $0.69 /hr, metered by the millisecond"; the separate /inference page ("Serverless inference for any model") describes the same endpoint product, so it is not a second row.

Sold within

Pricing page · 16 Nov 2026

Entitlement read 2026-11-16 with curl; page quote: "Concurrent vCPUs is the total CPU running at once across VMs, sandboxes, and serverless apps".

Name

Company disclosure · 16 Nov 2026

Product page text fetched 2026-11-16 with curl; page quote: "Run your model on RTX 4090 and RTX 5090 GPUs that start when a request arrives and shut down after traffic stops.". The page says "GPU time starts at $0.69 /hr, metered by the millisecond"; the separate /inference page ("Serverless inference for any model") describes the same endpoint product, so it is not a second row.

URL

Company disclosure · 16 Nov 2026

Product page text fetched 2026-11-16 with curl; page quote: "Run your model on RTX 4090 and RTX 5090 GPUs that start when a request arrives and shut down after traffic stops.". The page says "GPU time starts at $0.69 /hr, metered by the millisecond"; the separate /inference page ("Serverless inference for any model") describes the same endpoint product, so it is not a second row.

Also from Beam

Open Beam in the directory

Something wrong here? Send a correction — quote this product id: beam-serverless-gpu.