kitio.aiAdded Oct 28, 2026
Cerebrium
Freemium · $100/moServerless GPUs for voice agents, LLMs and video models
What is Cerebrium?
Cerebrium runs AI workloads such as voice agents, LLMs and image or video models on serverless GPUs billed by the second. It targets teams that need fast cold starts and autoscaling without managing Kubernetes.
You deploy your own code in containers, and memory and GPU snapshotting restores them in a few seconds when traffic arrives. Hardware ranges from CPU-only and T4 up to A100, H100, H200 and B200 GPUs across regions in the US, Europe and India.
Compute, memory and storage are charged per second on top of a platform plan, with the first 100 GB of storage free. The paid tier adds unlimited apps and seats, custom domains, a dedicated Slack channel and optional ML engineering help, and enterprise deals can reserve capacity.
Features
- Per-second GPU billing
- Fast cold starts with snapshots
- Autoscaling
- B200 to T4 GPUs
- Custom domains
- Observability
Tags
Releases
+ Submit new release- Listed
Cerebrium was added to the directory.
Organization
Cerebrium
Listed by kitio.ai · unverified owner
Pricing
- Pricing model
- Freemium
- Paid options from
- $100/mo
- Billing frequency
- Monthly
Cerebrium plans
Hobby
Free
- Plus compute usage
- 3 user seats
- Community support
Standard
$100/mo
- Plus compute usage
- Unlimited seats and apps
- Custom domains
- Dedicated Slack support
Enterprise
Custom
- Tailored pricing
- Reserved capacity options
Prices in USD, checked Oct 29, 2026. Plans change often, so confirm on the Cerebrium site.Official pricing
Sponsored
Your AI tool here
A Featured launch puts your tool in this spot and on the home page for a week.
Top Cerebrium alternatives
View allModels & APIs
Production inference platform for custom and open-source models
Models & APIs
The main hub for open models, datasets and AI demo apps
Models & APIs
NVIDIA's optimized model containers and hosted API catalog
Models & APIs
Serverless inference for open models on Cloudflare's global network
Compare Cerebrium
If you liked Cerebrium
Very fast inference API for open models on custom LPU chips
FreemiumModels & APIsMade in United StatesOn-demand GPU pods and serverless endpoints for AI workloads
PaidModels & APIsMade in United StatesRun and deploy open-source AI models through a simple API
PaidModels & APIsMade in United StatesFast inference and fine-tuning platform for open models
PaidModels & APIsMade in United StatesServerless cloud for running Python and GPU workloads on demand
Freemium · $250/moModels & APIsMade in United States
Pros and cons
Pros
- Pay only for seconds used
- Wide choice of GPU types
- Fast scale-up for real-time apps
Cons
- Cloud credits from AWS or GCP don't apply
- Paid plan adds $100 on top of compute
- Few regions available
Prompts & results
No prompts yet. Add your own prompt and output to help others use this tool.
Share a prompt
Reviews
Q&A
What is Cerebrium?
Cerebrium runs AI workloads such as voice agents, LLMs and image or video models on serverless GPUs billed by the second. It targets teams that need fast cold starts and autoscaling without managing Kubernetes.
Can I use AWS or GCP credits on Cerebrium?
No. Cerebrium bills separately, so cloud provider credits can't be applied to usage on the platform.
Which regions does Cerebrium run in?
Its site lists us-east-1, eu-west-2, eu-north-1 and ap-south-1.
How much does Cerebrium cost?
Cerebrium is freemium, with paid plans from $100/mo.
Is there a free version of Cerebrium?
Yes. Cerebrium has a free plan, and paid plans add more usage and features.
What are the main features of Cerebrium?
Per-second GPU billing, Fast cold starts with snapshots, Autoscaling, B200 to T4 GPUs, Custom domains, Observability.
Who makes Cerebrium?
Cerebrium is made by Cerebrium.
What are the best alternatives to Cerebrium?
Popular alternatives include Baseten, Hugging Face, NVIDIA NIM.