canonical: https://jentic.com/apis/lambdalabs.com/lambdalabs

# Lambdalabs Lambda Cloud API

The Lambda Cloud API launches and manages on-demand GPU compute instances for training and inference workloads. It lets you pick an instance type and region, launch, restart, and terminate instances, attach persistent filesystems, manage SSH keys and inbound firewall rules, browse available images, and open support tickets with attachments. Requests are authenticated with a Lambda Cloud API key.

## For AI agents

Launch, list, restart, and terminate GPU instances, manage filesystems, SSH keys, and firewall rules, and open support tickets on Lambda Cloud.

## Scope

Does not handle billing, model training code, or data labeling. Use for provisioning and managing GPU cloud instances only.

## Capabilities

- Launch on-demand GPU instances from a chosen instance type and region
- List running instances and retrieve the details of one
- Restart or terminate instances when a job finishes
- Create and delete filesystems for persistent storage across instances
- Manage SSH keys and inbound firewall rules that control instance access
- Browse available images and open support tickets with attachments

## Use cases

### Spin up GPU compute for an AI agent job

An AI agent integration created through Jentic needs GPU capacity on demand. The Lambda Cloud API lists available instance types, launches an instance in a region, and terminates it when the run finishes, so the agent provisions exactly the compute a training or inference job needs and releases it afterwards.

Example prompt: List instance types, launch one matching the job, then terminate it after the run completes

### Persistent storage across instances

Training runs need data that outlives a single instance. The Lambda Cloud API creates filesystems and lists them, so an automation can attach persistent storage to new instances and keep datasets and checkpoints available across launches without re-copying them each time.

Example prompt: Create a filesystem, then launch an instance that mounts it for a training run

### Lock down instance access

GPU instances exposed to the internet need controlled access. The Lambda Cloud API manages SSH keys and replaces inbound firewall rules, letting an agent register the right key and restrict inbound traffic to known addresses before a workload starts.

Example prompt: Add an SSH key and replace the inbound firewall rules to allow only a specified address

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | `/api/v1/instances` | List running instances |
| POST | `/api/v1/instance-operations/launch` | Launch instances |
| POST | `/api/v1/instance-operations/terminate` | Terminate instances |
| GET | `/api/v1/instance-types` | List available instance types |
| POST | `/api/v1/filesystems` | Create a filesystem |
| GET | `/api/v1/images` | List available images |
| POST | `/api/v1/ssh-keys` | Add an SSH key |
| GET | `/api/v1/firewall-rules` | List inbound firewall rules |

## Key resources

- **Instances** — Launch, list, restart, terminate, and retrieve GPU compute instances
- **Instance types** — List the available instance types and regions to launch into
- **Filesystems** — Create, list, and delete persistent filesystems
- **SSH keys** — List, add, and delete SSH keys for instance access
- **Firewall rules** — List and replace inbound firewall rules and rulesets
- **Tickets** — Create, list, and update support tickets with attachments

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 64 / 100
- **Maturity:** AI-Aware
- **Dimensions:**
  - Foundational Compliance: 99 / 100
  - Developer Experience & Jentic Compatibility: 67 / 100
  - AI-Readiness & Agent Experience: 53 / 100
  - Agent Usability: 94 / 100
  - Security: 35 / 100
  - AI Discoverability: 74 / 100
- **View full report:** https://jentic.com/apis/lambdalabs.com/lambdalabs/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** Wiring Lambda Cloud by hand means handling its API key over bearer or basic auth and registering SSH keys and firewall rules before you can launch an instance. Through Jentic you install once, import Lambda Cloud from the API Directory, store the key once, and your agent calls it.
- **Permission scoping:** Lambda Cloud puts the instance id in the URL path for the detail and update operations, so a rule can limit your agent to specific instances for reads and updates. You choose the operations it may call, so destructive ones like launching or terminating instances are not included unless you add them.
- **Credential handling:** Your Lambda Cloud API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'launch a GPU instance' or 'list instance types', and Jentic returns the matching Lambda Cloud operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

## Related APIs

- **DigitalOcean** — DigitalOcean provisions general-purpose and GPU droplets; Lambda Cloud is focused on GPU instances for AI workloads.
- **Linode** — Linode offers cloud compute instances and storage; Lambda Cloud specializes in GPU instances with persistent filesystems.
- **Replicate** — Replicate runs hosted models on managed GPUs; pair it with Lambda Cloud when you want to run your own code on raw instances.

## FAQ

### What authentication does the Lambda Cloud API use?

The Lambda Cloud API authenticates with your Lambda Cloud API key, sent either as a bearer token in the Authorization header or via HTTP basic auth, per its OpenAPI spec. Through Jentic the key is stored encrypted by your own Jentic One instance and injected at call time, so it never reaches the agent's prompt or logs.

### Can I launch and terminate GPU instances with the Lambda Cloud API?

Yes. The launch operation starts instances of a chosen type and region, and the terminate operation releases them, so an agent can provision GPU compute for a job and tear it down when the run finishes.

### What are the rate limits for the Lambda Cloud API?

The Lambda Cloud API spec notes that requests are generally limited to one request per second, with the instance launch operation limited to one request per 12 seconds. Design agent flows to respect those limits and back off on throttling.

### How do I launch a GPU instance through Jentic?

Install Jentic One on your own infrastructure with `curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh`, then import the Lambda Cloud API from the API Directory and search for 'launch a GPU instance'. Store the API key once and your agent can provision and release instances. To run it on your own infrastructure, install Jentic One from its GitHub repo.

### Is there a Lambda Cloud MCP server?

You don't need an MCP server to give your agent the Lambda Cloud API. Jentic connects it directly from the API Directory: import it, store your key once, and your agent calls it, with operations discovered on demand instead of loaded into the agent's context.

### Can I limit what my agent is allowed to do with the Lambda Cloud API?

Yes. The instance id sits in the URL path for the detail and update operations, so a rule can scope the agent to reading and updating specific instances, while destructive operations like launching or terminating stay out of reach unless you add them. Every action it takes is recorded in an audit log.