AI Tools
RunPod Serverless
RunPod Serverless: preços, capacidades, atualizações e alternativas, lidos nas páginas do próprio desenvolvedor.

Em inglês
Key facts
- Price
- Paid plans: see pricing Read on runpod.io
- Free plan
- No Read on runpod.io
- Platforms
- Web · API
- API
- Yes Read on runpod.io
- Open source
- Not presented as open source
- Last update
- 15 Nov 2026 Read on docs.runpod.io
- Prices checked
- 11 Nov 2026 Read on runpod.io
Developers and teams deploy their own containerized models as API endpoints without managing GPU servers. Workers can scale with requests, including scaling Flex workers to zero; the service also supports batch jobs and persistent storage options.
GPU compute is metered by the second, with rates varying by GPU. Storage may cost extra, and batch jobs are for work that can wait because start times are not guaranteed. The product page describes inference and other compute workloads, not a hosted general-purpose chatbot.
- Category
- Developer Tools
- Read from the maker’s pages
- 2026-11-18
Company
Features
- Deploy custom containerized models as API endpoints.
- Scale Flex workers with demand, including scaling to zero.
- Use FlashBoot for advertised sub-200ms cold starts.
- Choose GPU worker types with per-second compute billing.
- Run batch jobs on dedicated workers alongside live endpoint traffic.
- Attach storage for model weights and other persistent files.
- Deploy and roll back endpoints through GitHub integration.
Use cases
- Serve image, text, or audio model inference through an API.
- Process large embedding or inference jobs as batch work.
- Scale GPU workers for an application with variable request traffic.
- Deploy a private model endpoint for an internal chatbot.
- Run model workloads without managing GPU servers.
Pricing
Paid plans: see pricing
- Prices checked
- 2026-11-11
B300
$ 11.99 /hr
- $ 11.99 /hr
- 288 GB
- Maximum throughput for big models.
B200
$ 10.79 /hr
- $ 10.79 /hr
- 180 GB
- Maximum throughput for big models.
H200
$ 6.99 /hr
- $ 6.99 /hr
- 141 GB
- Extreme throughput for big models.
RTX 6000 Pro
$ 3.49 /hr
- $ 3.49 /hr
- 96 GB
- PRO
- High throughput for large model inference workloads.
H100
$ 5.39 /hr
- $ 5.39 /hr
- 80 GB
- PRO
- Extreme throughput for big models.
A100
$ 2.72 /hr
- $ 2.72 /hr
- 80 GB
- High throughput GPU, yet still very cost-effective.
L40, L40S, 6000 Ada, MIG 48GB
$ 1.75 /hr
- $ 1.75 /hr
- 48 GB
- PRO
- Extreme inference throughput on LLMs like Llama 3 7B.
A6000, A40
$ 1.22 /hr
- $ 1.22 /hr
- 48 GB
- A cost-effective option for running big models.
5090
$ 1.79 /hr
- $ 1.79 /hr
- 32 GB
- PRO
- Extreme throughput for small-to-medium models.
RTX PRO 4500 Blackwell
$ 1.15 /hr
- $ 1.15 /hr
- 32 GB
- Cost-effective Blackwell inference for 32GB workloads.
Capabilities
- Command line — “CLI & SDKs. Deploy and manage directly from your terminal.” source
- API — “Serverless GPU endpoints run containerized inference workloads behind an API and scale workers based on demand.” source
- Official SDKs — “CLI & SDKs. Deploy and manage directly from your terminal.” source
- Runs models for you — “Deploy any containerized model as an autoscaling GPU endpoint without managing servers.” source
- Builds agents and workflows — “Build intelligent agent-based systems and workflows.” source
- Traces and evaluates — “Runpod offers a comprehensive monitoring dashboard with real-time logging and distributed tracing for your serverless functions.” source
Latest updates
- Breaking REST API v1 retired
Migrate your integrations to REST API v2 to restore access.
- New Release Scoped API keys
You can now choose each API key’s access to Runpod services separately, including per-scope access for the REST API v2.
- Terraform provider 2.0.0
Built on the REST API v2 and adds support for Instant Clusters, global volumes, CPU Pods and endpoints, FlashBoot tiers, and v2 billing.
- Terraform provider 1.x to 2.0.0 migration
Upgrading from 1.x requires configuration and state changes; 2.0.0 makes no REST v1 calls.
- New Release Terraform provider 2.0.0
Built on the REST API v2, with support for Instant Clusters, global volumes, CPU Pods and endpoints, FlashBoot tiers, and v2 billing.
Where it works
Web · API
Alternatives to RunPod Serverless
- Modal
Modal runs containerized applications on serverless cloud infrastructure, including GPU inference, training, batch jobs, and isolated code sandboxes.
- RunPod Pods
RunPod Pods provide dedicated cloud GPUs for AI development, training, fine-tuning, inference, and long-running jobs.
- Krutrim Cloud
Krutrim Cloud provides CPU and GPU computing, storage, networking, Kubernetes, and AI services for deploying applications and AI workloads.
- Together AI
Together AI provides APIs and infrastructure for running, fine-tuning, and training open AI models.
- CoreWeave Cloud
Cloud infrastructure for training and serving AI, with GPU and CPU compute, storage, networking, and managed Kubernetes.
- IO Cloud
IO Cloud provides on-demand GPU clusters for AI workloads through io.net’s cloud platform.
More from RunPod
- RunPod Pods
RunPod Pods provide dedicated cloud GPUs for AI development, training, fine-tuning, inference, and long-running jobs.
Questions about RunPod Serverless
Is RunPod Serverless free?
No. The maker’s pricing page lists no free plan for RunPod Serverless. Prices checked on 11 Nov 2026.
How much does RunPod Serverless cost?
The plans on the maker’s pricing page: B300 ($ 11.99 /hr), B200 ($ 10.79 /hr), H200 ($ 6.99 /hr), RTX 6000 Pro ($ 3.49 /hr), H100 ($ 5.39 /hr), A100 ($ 2.72 /hr), L40, L40S, 6000 Ada, MIG 48GB ($ 1.75 /hr), A6000, A40 ($ 1.22 /hr), 5090 ($ 1.79 /hr) and RTX PRO 4500 Blackwell ($ 1.15 /hr). Prices checked on 11 Nov 2026.
Does RunPod Serverless have an API?
Yes. The maker’s pages state that RunPod Serverless has an API. Read on the maker’s pages on 18 Nov 2026.
Which platforms does RunPod Serverless run on?
RunPod Serverless runs on Web and API, as its maker’s pages state. Read on the maker’s pages on 18 Nov 2026.
Is RunPod Serverless open source?
The maker’s pages do not present RunPod Serverless as open source. Read on the maker’s pages on 18 Nov 2026.
Is RunPod Serverless still available?
Yes, as of 18 Nov 2026: its maker’s pages were read that day and carried no notice that RunPod Serverless has ended.
What changed recently in RunPod Serverless?
The latest entry in its maker’s changelog: 15 Nov 2026, “Breaking REST API v1 retired”. Before it: 12 Nov 2026, “New Release Scoped API keys”. Changelog checked on 19 Nov 2026.
Sources
- https://www.runpod.io/pricing — Official pricing page read on 2026-11-18; lists Serverless GPU rates and per-second billing.
- https://www.runpod.io/product/serverless — Product page describes containerized inference endpoints, autoscaling, API access, storage, and deployment capabilities.
- https://www.runpod.io/blog/batch-inference-is-now-generally-available-on-runpod — Runpod announcement dated 2026-10-28 describes Batch Inference general availability.
Not confirmed
- Latest version
- The product pages read do not identify a Runpod Serverless version or model name.