Beam
Serverless GPUs, sandboxes, and on-demand machines
Founded in 2021, Beam is a US platform for running AI workloads on GPUs without managing servers or Dockerfiles. Workloads are defined from a Python, TypeScript or Go SDK and run as serverless endpoints, task queues, or sandboxes for untrusted code.
Memory snapshots and GPU checkpoint restore give sub-second cold starts, across 30+ regions in the US, Europe and Asia. Beam also sells flat-rate on-demand machines and reserved InfiniBand clusters, or will run the same workloads in your own AWS, GCP or Azure account.
Example customers include Coke, Ogilvy, EdgeImpulse, Magellan AI, Geospy, Frase. More here.
Beam Homepage
What's good about Beam
- Pay-per-millisecond billing, only while code is running
- Sub-second cold starts from memory snapshots
- No egress or bandwidth fees on any plan
- Runs in your own AWS, GCP or Azure account
- Recurring monthly credit on every plan
Beam pricing examples
Beam offers three billing models:
- Serverless: billed by the millisecond, with GPU, CPU and RAM priced separately.
- On-demand: one flat hourly rate per machine, including vCPU, RAM and NVMe.
- Clusters: reserved monthly or yearly, quoted per customer.
Below are some example configurations and their estimated costs:
| Example configuration | Estimated cost |
|---|---|
| Block Storage | $2.10 /mo 100 GB beyond free allowance · First 1 TB included |
| Free egress allowance | Free and unlimited |
Beam GPUs
How this list works
Order
- By GPU model: one card per GPU, the most looked-up models first, then the most widely offered ones.
- By configuration: every configuration we track. What you can rent now comes first, and within that one row per GPU model before any repeats, so the opening rows cover the range rather than one model's sizes.
The price on a card
Each card leads with a per GPU per hour rate. We prefer configurations the provider reports as in stock, on-demand rates before reserved or spot, then take the lowest per-GPU rate among those.
Search
We match the GPU model, form factor, billing type, availability and instance name. Matching is partial and case-insensitive. Everyday words work too, so "interruptible" finds spot and "sold out" finds out of stock.
Transparency and funding
- Ads and sponsors: Paid placements sit at the top of the list and are always labeled as sponsored content. Sponsorship never influences the organic ranking itself.
- Affiliates: Affiliate links are marked. We may earn a commission if you click them, but commissions never affect the order.
- Prices: Shown in USD, converted at daily reference rates where a provider publishes in another currency. A month means 720 hours. How we estimate costs.
No GPU models match .
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Verify before provisioning. More on how we price. See Beam's pricing.
Which services does Beam offer
Here are some of the services that Beam offers:
Alternatives to Beam
Compare Beam against other cloud providers:
-
Replicate
Replicate offers serverless GPU computing with a focus on running pre-built AI models.
-
Runpod
Runpod provides GPU instances with serverless and dedicated options across multiple regions.
-
Together AI
Together AI specializes in inference for open-source language models.
-
Cerebrium
Cerebrium is a serverless GPU platform for deploying machine learning applications.
-
Fal.ai
fal is a serverless platform specialising in generative media model inference.
Our data for Beam was last updated on Sept. 11, 2026.