Modal
Bursty or unpredictable GPU inference that benefits from scale-to-zero and per-second billing
T4 16 GB · from:
$0.59≈ €0.527 (estimated conversion from $0.59)
Managed AI hosting compared: serverless GPUs, endpoints and apps
We may earn commissions when you shop through the links below. Learn more
Let the provider handle servers, scaling and updates. Compare managed platforms on verified GPU prices, billing increments, regions and what you still manage yourself.
Managed AI hosting providers
What managed means here
- The provider runs the runtime
Servers, drivers, scaling and restarts are handled for you; you deploy code, containers or models.
- You still own the rest
Your code, data, API keys and costs stay your responsibility, including limits on runaway usage.
- The trade-off
Less control over the machine and platform-specific limits, in exchange for less operations work.
Dify and Flowise: what their official docs say
Dify's Docker Compose deployment needs at least 2 CPU cores and 4 GiB of RAM, and Docker Compose 2.24.0 or newer.
A self-hosted Dify stack starts its own PostgreSQL database, Redis, a Weaviate vector database, Nginx and code sandboxes next to the Dify services, so size memory and disk for all of them.
Flowise's deployment docs list managed platforms such as Render and Railway next to AWS, Azure, Google Cloud, DigitalOcean and Kubernetes with Helm.
Tools for this decision
- GPU Cost CalculatorEstimate monthly GPU spend from verified hourly prices, usage hours, instances and storage.
- VRAM / Model FitEstimate the GPU memory a model needs for inference or fine-tuning, and which GPUs fit.
- Agent Hosting SelectorAnswer a few questions about your agent and get a best-fit hosting category and provider shortlist.
How we compare
Prices per GPU-hour.
GPU prices are USD per single GPU per hour from the provider's own pricing page or API. Multi-GPU nodes are divided by their GPU count, and a price only rented as a node says so (“rented as 8 GPUs”). Rows lead with the cheapest pay-as-you-go price (on-demand, secure, community or serverless). A spot, reserved or marketplace price leads only when the provider has no pay-as-you-go price here, or when you filter by that billing; it is labelled with its price type and listed in its own group after pay-as-you-go prices, never ranked against them. Prices exclude VAT and sales tax.Not verified means not verified.
When no official source confirms a value we show “Not verified” instead of an estimate. Where a provider's terms do not allow its prices to be republished, we link to its live pricing instead.Freshness on every row.
Updated today: checked on the current day (UTC). Fresh: up to 7 days. Stale: up to 30 days. Needs refresh: older than that.No paid placement.
Results only change order with the sort you pick. Affiliate relationships never affect order, filters or which providers appear.
Filters match a provider when one of its listed offers meets every GPU, VRAM, region, price and billing-increment condition together. Spot, reserved and serverless billing list each provider that offers that pricing and show only prices on that basis, or “not verified” when we have none. Read the full methodology and our affiliate disclosure.
Frequently asked questions
What is the difference between managed and self-managed hosting?
With managed hosting the provider operates the runtime: servers, updates, scaling and restarts. With self-managed hosting you get a machine or container with root or SSH access and run everything yourself.
How are managed AI platforms billed?
Mostly by usage, such as seconds of GPU or CPU time, sometimes on top of a monthly platform plan. The results show the per-GPU-hour rate; platform plans and other fees are listed on each provider page.
Where does the model run in Dify or Flowise?
Wherever you point it. Both builders call a model through an API key for a hosted model, or through your own endpoint, for example Ollama or vLLM on a GPU server. The app platform itself mainly needs CPU, memory and storage.
Why do some values say “Not verified”?
We only show a value when an official provider source confirms it. If we could not confirm a price or feature, the row says “Not verified” instead of showing an estimate, and the provider page lists the sources we did check.