Prime Intellect

https://docs.primeintellect.ai/
Admin
Prime Intellect provides comprehensive documentation for its AI platform, offering resources and...
Tokens:241,375
Snippets:1,556
Trust Score:9.3
Update:1 month ago
Show doc for...
Tokens:
Raw
### Lab setup output example

Source: https://docs.primeintellect.ai/guides/rl-training

Example output and quick command reference displayed after running the lab setup command.

```text
~/dev/demo prime lab setup
Supported coding agents: codex, claude, cursor, opencode, amp
Primary coding agent [codex]:
Using multiple coding agents? [y/N]:
No pyproject.toml found, initializing uv project...
Running: uv init
Initialized project `demo`
Running: uv add verifiers
...
... # install + download outputs omitted
...
[................................................................................] 1371 / 1371
Downloaded configs/rl/wordle.toml from https://github.com/primeintellect-ai/verifiers

+------------------------------------------ get started -------------------------------------------+
|                                                                                                  |
|  idea -> environment -> eval -> training                                                         |
|                                                                                                  |
|  +-------------------------------- ask codex ---------------------------------+                  |
|  |                                                                            |                  |
|  | I want to train a model for <my task domain>. Propose an initial           |                  |
|  | environment scaffold including relevant tools, generate a small            |                  |
|  | synthetic dataset, run a quick eval baseline, inspect the results,         |                  |
|  | and decide how to iterate on refining the implementation.                  |                  |
|  |                                                                            |                  |
|  +----------------------------------------------------------------------------+                  |
|                                                                                                  |
|  +-------------------- quick commands --------------------+                                      |
|  |                                                        |                                      |
|  | $ prime env init my-env                                |                                      |
|  | $ prime eval run my-env -m gpt-5-nano -n 5             |                                      |
|  | $ prime eval tui                                       |                                      |
|  | $ prime train run configs/rl/wiki-search.toml             |                                      |
|  | $ prime gepa run my-env -m gpt-5-nano                  |                                      |
|  |                                                        |                                      |
|  +--------------------------------------------------------+                                      |
|                                                                                                  |
+--------------------------------------------------------------------------------------------------+
```

--------------------------------

### Install and Setup Verifiers

Source: https://docs.primeintellect.ai/verifiers/v1/getting_started

Use these commands to install uv, clone the repository, and synchronize dependencies.

```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
git clone https://github.com/PrimeIntellect-ai/verifiers.git
cd verifiers
uv sync
```

--------------------------------

### Install Training Environment

Source: https://docs.primeintellect.ai/hosted-training/end-to-end-run

Downloads and installs a specific environment for training.

```bash
prime env install primeintellect/alphabet-sort
```

--------------------------------

### Run Browser Examples

Source: https://docs.primeintellect.ai/guides/browser-environments

Execute the browser-dom or browser-cua example tasks.

```bash
# DOM mode
prime eval run browser-dom-example -m openai/gpt-4.1-mini

# CUA mode (pre-built image, fastest and recommended)
prime eval run browser-cua-example -m openai/gpt-4.1-mini
```

--------------------------------

### Install an environment

Source: https://docs.primeintellect.ai/tutorials-environments/install

Install an environment from the Environments Hub using the CLI.

```bash
prime env install <owner>/<environment-name>
```

--------------------------------

### Alternative installation methods

Source: https://docs.primeintellect.ai/tutorials-environments/install

Install environments directly using pip or uv via the wheel URL.

```bash
# Direct pip installation from wheel URL
pip install https://hub.primeintellect.ai/will/wordle/@latest/wordle-0.1.4-py2.py3-none-any.whl

# UV pip installation
uv pip install https://hub.primeintellect.ai/will/wordle/@latest/wordle-0.1.4-py2.py3-none-any.whl

# Add to existing project with UV
uv add wordle@https://hub.primeintellect.ai/will/wordle/@latest/wordle-0.1.4-py2.py3-none-any.whl
```

--------------------------------

### Install prime-rl

Source: https://docs.primeintellect.ai/prime-rl/overview

Use the provided installation script to clone the repository, initialize submodules, and set up the environment using uv.

```bash
curl -sSL https://raw.githubusercontent.com/PrimeIntellect-ai/prime-rl/main/scripts/install.sh | bash
```

--------------------------------

### Install CLI and Authenticate

Source: https://docs.primeintellect.ai/hosted-training/getting-started

Installs the uv package manager and the Prime CLI, then initiates the login process.

```bash
# install uv (if not already installed)
curl -LsSf https://astral.sh/uv/install.sh | sh
# install the prime CLI
uv tool install -U prime
# log in
prime login
```

--------------------------------

### Install prime CLI

Source: https://docs.primeintellect.ai/guides/rl-training

Installs the prime command-line interface tool using uv.

```bash
uv tool install prime
```

--------------------------------

### Install and Evaluate the Environment

Source: https://docs.primeintellect.ai/guides/browser-environments

Use uv to install the local environment package and run the evaluation using the prime CLI.

```bash
uv pip install -e ./environments/your_env
prime eval run your-env -m openai/gpt-4.1-mini
```

--------------------------------

### Test Environment Prompt

Source: https://docs.primeintellect.ai/guides/rl-training

Example prompt used to verify environment performance with GPT-4.1 models before starting a training run.

```text
test with GPT-4.1 models, make sure we're seeing proper rollouts that succeed with non-zero scores. those models should be able to solve; if they can't we have a bug somewhere or need clearer instructions in the env. Once you feel good about the env, make a RL config and start a training run with bs 128, 8 rollouts per group, for the 30b instruct qwen model.
```

--------------------------------

### Setup Workspace and Authenticate

Source: https://docs.primeintellect.ai/hosted-training/end-to-end-run

Initializes the development environment and authenticates the CLI.

```bash
# Install and authenticate the CLI
uv tool install prime
prime login

# Set up a workspace
mkdir ~/dev/my-lab && cd ~/dev/my-lab
prime lab setup
```

--------------------------------

### Install environment versions

Source: https://docs.primeintellect.ai/tutorials-environments/install

Install the latest version or a specific version of an environment.

```bash
# Install latest version
prime env install will/wordle

# Install specific version
prime env install will/wordle@0.1.3
```

--------------------------------

### Install and Test Environment

Source: https://docs.primeintellect.ai/tutorials-environments/create

Commands for installing the environment locally and running evaluations using the verifiers library.

```bash
uv pip install -e .
```

```bash
uv run vf-eval my-environment
```

--------------------------------

### Verify and install environments

Source: https://docs.primeintellect.ai/hosted-training/troubleshooting

Commands to check environment information and install specific versions from the Hub.

```bash
# Check the exact ID
prime env info owner/environment-name

# Install a specific version
prime env install owner/environment-name@latest
```

--------------------------------

### Install Browser Dependencies

Source: https://docs.primeintellect.ai/guides/browser-environments

Install the necessary verifiers package with browser support.

```bash
# Install verifiers with browser support
uv pip install "verifiers[browser]"
```

--------------------------------

### Install Prime Intellect CLI

Source: https://docs.primeintellect.ai/cli-reference/introduction

Installs the CLI tool using uv or pip.

```bash
uv tool install prime
```

```bash
pip install prime
```

--------------------------------

### Configure Sandbox Start Command

Source: https://docs.primeintellect.ai/sandboxes/sdk

Defines the process to run upon sandbox boot. The process is one-shot; the sandbox stops when the process exits.

```python
from prime_sandboxes import StartCommand

sandbox = await sandboxes.create(
    CreateSandboxRequest(
        name="api-server",
        docker_image="python:3.11-slim",
        start_command=StartCommand(executable="python", args=["serve.py", "--port", "8000"]),
    )
)
```

```python
start_command=StartCommand(executable="/bin/sh", args=["-c", "python serve.py --port 8000"])
```

--------------------------------

### Install uv

Source: https://docs.primeintellect.ai/cli-reference/introduction

Installs the uv package manager required for CLI installation.

```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
```

--------------------------------

### Run a Quick RL Example

Source: https://docs.primeintellect.ai/prime-rl/overview

Execute a training run using the rl entrypoint and a specified TOML configuration file.

```bash
uv run rl @ examples/basic/reverse-text/rl.toml
```

--------------------------------

### Install or Update Prime CLI

Source: https://docs.primeintellect.ai/hosted-training/troubleshooting

Commands to install, upgrade, or verify the Prime CLI installation.

```bash
# Install or reinstall
uv tool install prime

# If already installed, upgrade
uv tool install -U prime

# Verify
prime --version   # Should be >= 0.5.15
```

--------------------------------

### Initialize Workspace

Source: https://docs.primeintellect.ai/hosted-training/getting-started

Sets up the local directory with necessary configuration files and environment structures.

```bash
# ~/dev/my-lab
prime lab setup
```

--------------------------------

### Initialize New Environment

Source: https://docs.primeintellect.ai/tutorials-environments/create

Creates a starter template for a Python module including a README, pyproject.toml, and stub code.

```bash
prime env init <your-env-name>
```

--------------------------------

### Port Mapping Example

Source: https://docs.primeintellect.ai/api-reference/pods/create-pod

Example values for internal and external port mappings.

```yaml
internal:
  type: string
  title: Internal
  examples:
    - '8080'
external:
  type: string
  title: External
  examples:
    - '1234'
```

--------------------------------

### Initialize bb_demo environment

Source: https://docs.primeintellect.ai/guides/browser-environments

Load the bb_demo environment with optional screenshot saving and click marker visualization enabled.

```python
from bb_demo import load_environment

env = load_environment(
    save_screenshots=True,
    screenshot_dir="./my_screenshots",
    mark_clicks=True,
    click_marker_radius=15,
)
```

--------------------------------

### prime tunnel start

Source: https://docs.primeintellect.ai/sandboxes/tunnel

CLI command to start a tunnel for a local service.

```APIDOC
## CLI: prime tunnel start

### Description
Starts a tunnel to expose a local service to the internet.

### Usage
`prime tunnel start --port <port> [--auth <username>]`

### Parameters
- **--port** (int) - Required - The local port to expose.
- **--auth** (str) - Optional - Enables basic authentication with the specified username.
```

--------------------------------

### Start a local HTTP server

Source: https://docs.primeintellect.ai/sandboxes/tunnel

Create a simple HTML file and serve it locally on port 8000.

```bash
# Create a simple HTML file
echo '<h1>Hello from Prime Tunnel!</h1>' > index.html

# Start Python's built-in HTTP server on port 8000
python -m http.server 8000
```

--------------------------------

### Instance Provisioning Response

Source: https://docs.primeintellect.ai/api-reference/provision-gpu

Example JSON response returned upon successful instance provisioning, including resource and disk details.

```json
{
  "id": "11de9260fb52451085f3936c20b0ffdb",
  "userId": "clvb0itli0000oxevorvgrpfn",
  "teamId": null,
  "walletId": "clvb0iv0c0004oxev5mnt1xiv",
  "name": "Training instance with data",
  "providerType": "hyperstack",
  "status": "PROVISIONING",
  "installationStatus": "PENDING",
  "installationFailure": null,
  "installationProgress": null,
  "createdAt": "2025-11-07T23:36:36.391000",
  "updatedAt": "2025-11-07T23:36:36.418000",
  "terminatedAt": null,
  "gpuName": "H100_80GB",
  "gpuCount": 1,
  "socket": "PCIe",
  "priceHr": 1.9,
  "stoppedPriceHr": null,
  "provisioningPriceHr": 0.0,
  "environmentType": "ubuntu_22_cuda_12",
  "customTemplateId": null,
  "clusterId": null,
  "primePortMapping": [
    {
      "internal": "22",
      "external": "22",
      "protocol": "TCP",
      "usedBy": "SSH",
      "description": "SSH access"
    },
    {
      "internal": "*",
      "external": "*",
      "protocol": "TCP",
      "usedBy": null,
      "description": "All other ports"
    }
  ],
  "sshConnection": null,
  "ip": null,
  "resources": {
    "memory": 180,
    "disk": 1250,
    "sharedDisk": null,
    "vcpus": 28
  },
  "attachedResources": [
    {
      "resourceType": "DISK",
      "id": "clhxy6aw80000j8080gdf8kqv",
      "status": "UNATTACHED",
      "isDetachable": true,
      "mountPath": "/data",
      "resourcePath": "/dev/vdc",
      "size": 500,
      "isShared": false
    },
    {
      "resourceType": "DISK",
      "id": "ephemeral",
      "status": "ATTACHED",
      "isDetachable": false,
      "mountPath": "/ephemeral",
      "resourcePath": "/dev/vdb",
      "size": 750,
      "isShared": false
    }
  ],
  "isSpot": null,
  "autoRestart": null
}
```

--------------------------------

### Launch Training Run

Source: https://docs.primeintellect.ai/hosted-training/getting-started

Starts a training job using a specified configuration file from the local workspace.

```bash
prime train run configs/rl/alphabet-sort.toml
```

--------------------------------

### Start Training Run

Source: https://docs.primeintellect.ai/guides/rl-training

Command to initiate a training run using a specific configuration file.

```bash
prime train run configs/rl/calendar-scheduling.toml
```

--------------------------------

### Install missing verifiers module

Source: https://docs.primeintellect.ai/hosted-training/troubleshooting

Use these commands to install the missing verifiers library or a specific environment.

```bash
uv add verifiers
```

```bash
prime env install my-env
```

--------------------------------

### Initialize Lab workspace

Source: https://docs.primeintellect.ai/guides/rl-training

Prepares the current directory as a Lab workspace by bootstrapping a Python project and downloading necessary configurations.

```bash
prime lab setup
```

--------------------------------

### OpenAPI Specification for Get Whoami

Source: https://docs.primeintellect.ai/api-reference/user/get-whoami

Defines the GET /api/v1/user/whoami endpoint, including security requirements and response schemas.

```yaml
openapi: 3.1.0
info:
  title: PI API
  version: 0.1.0
servers:
  - url: https://api.primeintellect.ai
security: []
paths:
  /api/v1/user/whoami:
    get:
      tags:
        - user
      summary: Get Whoami
      operationId: get_whoami_api_v1_user_whoami_get
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/GenericResponse_WhoamiResponse_'
        '401':
          description: Authorization failed
        '422':
          description: Invalid request data
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
      security:
        - HTTPBearer: []
components:
  schemas:
    GenericResponse_WhoamiResponse_:
      properties:
        data:
          anyOf:
            - $ref: '#/components/schemas/WhoamiResponse'
            - type: 'null'
          description: Response data
        status:
          anyOf:
            - type: string
            - type: 'null'
          title: Status
          description: Response status
      type: object
      title: GenericResponse[WhoamiResponse]
    ErrorResponse:
      properties:
        errors:
          items:
            $ref: '#/components/schemas/ErrorDetail'
          type: array
          title: Errors
      type: object
      required:
        - errors
      title: ErrorResponse
    WhoamiResponse:
      properties:
        id:
          type: string
          title: Id
        role:
          $ref: '#/components/schemas/UserRoleEnum'
          default: USER
        createdAt:
          type: string
          format: date-time
          title: Createdat
        name:
          type: string
          title: Name
        slug:
          anyOf:
            - type: string
            - type: 'null'
          title: Slug
        email:
          type: string
          title: Email
        image:
          anyOf:
            - type: string
            - type: 'null'
          title: Image
        emailVerified:
          anyOf:
            - type: string
              format: date-time
            - type: 'null'
          title: Emailverified
        hasBetaAccess:
          type: boolean
          title: Hasbetaaccess
          default: false
        isBanned:
          type: boolean
          title: Isbanned
          default: false
        skipPrepay:
          type: boolean
          title: Skipprepay
          default: false
        blacklist:
          items:
            type: string
          type: array
          title: Blacklist
          default: []
        scope:
          additionalProperties:
            anyOf:
              - $ref: '#/components/schemas/PermissionType'
              - type: 'null'
          type: object
          title: Scope
          default: {}
      type: object
      required:
        - name
        - email
      title: WhoamiResponse
      description: Extended User response that includes token scope permissions.
    ErrorDetail:
      properties:
        param:
          type: string
          title: Param
        details:
          type: string
          title: Details
      type: object
      required:
        - param
        - details
      title: ErrorDetail
    UserRoleEnum:
      type: string
      enum:
        - ADMIN
        - MANAGER
        - USER
      title: UserRoleEnum
    PermissionType:
      properties:
        read:
          type: boolean
          title: Read
        write:
          type: boolean
          title: Write
      type: object
      required:
        - read
        - write
      title: PermissionType
  securitySchemes:
    HTTPBearer:
      type: http
      scheme: bearer
```

--------------------------------

### Configure OPSD Algorithm

Source: https://docs.primeintellect.ai/prime-rl/algorithms

Example configuration for the OPSD algorithm specifying the demo key.

```toml
[orchestrator.algo]
type = "opsd"
demo_key = "demonstration"
```

--------------------------------

### Launch a full fine-tuning run

Source: https://docs.primeintellect.ai/hosted-training/full-finetuning

Initiates a training run using a configuration file. A [deployment] block in the TOML file is required for auto-detection.

```bash
prime train run configs/full-ft.toml
```

--------------------------------

### OpenAPI Specification for Get Pods

Source: https://docs.primeintellect.ai/api-reference/pods/get-pods

The OpenAPI definition for the GET /api/v1/pods/ endpoint, detailing pagination parameters and response structures.

```yaml
openapi: 3.1.0
info:
  title: PI API
  version: 0.1.0
servers:
  - url: https://api.primeintellect.ai
security: []
paths:
  /api/v1/pods/:
    get:
      tags:
        - Pods
      summary: Get Pods
      operationId: get_pods_api_v1_pods__get
      parameters:
        - name: offset
          in: query
          required: false
          schema:
            type: integer
            default: 0
            title: Offset
        - name: limit
          in: query
          required: false
          schema:
            type: integer
            default: 100
            title: Limit
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: >-
                  #/components/schemas/GenericPaginationResponse_List_APIPodConfig__
        '401':
          description: Authorization failed
        '422':
          description: Invalid request data
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
      security:
        - HTTPBearer: []
components:
  schemas:
    GenericPaginationResponse_List_APIPodConfig__:
      properties:
        total_count:
          type: integer
          title: Total Count
          description: Total number of items available in the dataset
          default: 0
        offset:
          type: integer
          minimum: 0
          title: Offset
          description: Number of items to skip before starting to collect the result set
          default: 0
        limit:
          type: integer
          minimum: 0
          title: Limit
          description: Maximum number of items to return
          default: 100
        data:
          items:
            $ref: '#/components/schemas/APIPodConfig'
          type: array
          title: Data
        status:
          anyOf:
            - type: string
            - type: 'null'
          title: Status
          description: Response status
      type: object
      required:
        - data
      title: GenericPaginationResponse[List[APIPodConfig]]
    ErrorResponse:
      properties:
        errors:
          items:
            $ref: '#/components/schemas/ErrorDetail'
          type: array
          title: Errors
      type: object
      required:
        - errors
      title: ErrorResponse
    APIPodConfig:
      properties:
        id:
          type: string
          title: Id
          description: Unique identifier for the pod, generated as a UUID.
        userId:
          type: string
          title: Userid
          description: ID of the user associated with this pod, if applicable.
        teamId:
          anyOf:
            - type: string
            - type: 'null'
          title: Teamid
          description: ID of the team owning this pod, if applicable.
        walletId:
          anyOf:
            - type: string
            - type: 'null'
          title: Walletid
          description: >-
            ID of the wallet associated with this pod for billing or resource
            tracking.
        name:
          type: string
          title: Name
          description: Name of the pod.
        type:
          $ref: '#/components/schemas/PodTypeEnum'
          description: Type of the pod, based on PodTypeEnum.
          default: HOSTED
          deprecated: true
        providerType:
          $ref: '#/components/schemas/ProviderTypeEnum'
          description: Type of provider associated with the pod, based on ProviderTypeEnum.
        status:
          $ref: '#/components/schemas/PodStatusEnum'
          description: Current status of the pod, based on PodStatusEnum.
          default: PROVISIONING
        installationStatus:
          $ref: '#/components/schemas/InstallationStatusEnum'
          description: Installation status of the pod, based on InstallationStatusEnum.
          default: PENDING
        installationFailure:
          anyOf:
            - type: string
            - type: 'null'
          title: Installationfailure
          description: >-
            Details about any installation failures that occurred, if
            applicable.
        installationProgress:
          anyOf:
            - type: integer
            - type: 'null'
          title: Installationprogress
          description: Percentage of the installation process completed.
        createdAt:
          type: string
          format: date-time
          title: Createdat
          description: Timestamp when the pod was created.
        updatedAt:
          type: string
          format: date-time
          title: Updatedat
```

--------------------------------

### Manage Environments via CLI

Source: https://docs.primeintellect.ai/tutorials-environments/getting-started

Use these commands to list, inspect, install, or initialize reinforcement learning environments.

```bash
# List available environments
prime env list

# Get information about an environment
prime env info owner/environment-name

# Install an environment
prime env install owner/environment-name

# Create a new environment
prime env init my-new-environment
```

--------------------------------

### Basic Evaluation Workflow

Source: https://docs.primeintellect.ai/tutorials-environments/evaluating

Commands to list available environments, install a specific environment, and execute a local evaluation.

```bash
# List available environments
prime env list --owner primeintellect

# Install an environment
prime env install primeintellect/gsm8k@latest

# Run a basic evaluation
prime eval gsm8k
```

--------------------------------

### OpenAPI Specification for Get Disks Availability

Source: https://docs.primeintellect.ai/api-reference/availability/get-disks-availability

Defines the GET request structure, query parameters, and response schemas for disk availability.

```yaml
openapi: 3.1.0
info:
  title: PI API
  version: 0.1.0
servers:
  - url: https://api.primeintellect.ai
security: []
paths:
  /api/v1/availability/disks:
    get:
      tags:
        - Availability
      summary: Get Disks Availability
      operationId: get_disks_availability_api_v1_availability_disks_get
      parameters:
        - name: page
          in: query
          required: false
          schema:
            type: integer
            minimum: 1
            description: Page number (1-indexed)
            default: 1
            title: Page
          description: Page number (1-indexed)
        - name: page_size
          in: query
          required: false
          schema:
            type: integer
            maximum: 100
            minimum: 1
            description: Number of results per page
            default: 100
            title: Page Size
          description: Number of results per page
        - name: regions
          in: query
          required: false
          schema:
            anyOf:
              - type: array
                items:
                  $ref: '#/components/schemas/Regions'
              - type: 'null'
            description: List of regions to filter
            title: Regions
          description: List of regions to filter
        - name: data_center_id
          in: query
          required: false
          schema:
            anyOf:
              - type: string
              - type: 'null'
            description: Filter by data center ID
            title: Data Center Id
          description: Filter by data center ID
        - name: cloud_id
          in: query
          required: false
          schema:
            anyOf:
              - type: string
              - type: 'null'
            description: Filter by cloud ID
            title: Cloud Id
          description: Filter by cloud ID
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/DiskAvailabilityResponse'
        '401':
          description: Authorization failed
        '422':
          description: Invalid request data
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
      security:
        - HTTPBearer: []
components:
  schemas:
    Regions:
      type: string
      enum:
        - africa
        - asia_south
        - asia_northeast
        - australia
        - canada
        - eu_east
        - eu_north
        - eu_west
        - middle_east
        - south_america
        - united_states
      title: Regions
    DiskAvailabilityResponse:
      properties:
        items:
          items:
            $ref: '#/components/schemas/DiskAvailability'
          type: array
          title: Items
          description: List of available disk configurations
        totalCount:
          type: integer
          title: Totalcount
          description: Total number of matching disk configurations
      type: object
      required:
        - items
        - totalCount
      title: DiskAvailabilityResponse
    ErrorResponse:
      properties:
        errors:
          items:
            $ref: '#/components/schemas/ErrorDetail'
          type: array
          title: Errors
      type: object
      required:
        - errors
      title: ErrorResponse
    DiskAvailability:
      properties:
        cloudId:
          anyOf:
            - type: string
            - type: 'null'
          title: Cloudid
          description: Cloud ID, used for provisioning
        provider:
          $ref: '#/components/schemas/ProviderTypeEnum'
          description: Provider's ID
        dataCenter:
          anyOf:
            - type: string
            - type: 'null'
          title: Datacenter
          description: Location of the data center housing the disk.
        country:
          anyOf:
            - type: string
            - type: 'null'
          title: Country
          description: Country where the data center is located.
        region:
          anyOf:
            - $ref: '#/components/schemas/Regions'
            - type: 'null'
          description: Region where the disk is located.
        spec:
          $ref: '#/components/schemas/CustomSpecValue'
          description: Configurable shared disk size in GB and price per GB per hour.
          default: {}
        stockStatus:
          type: string
          enum:
            - Available
            - Low
            - Medium
            - High
            - Unavailable
          title: Stockstatus
```

--------------------------------

### OpenAPI Specification for Get Pods Status

Source: https://docs.primeintellect.ai/api-reference/pods/get-pods-status

Defines the GET /api/v1/pods/status endpoint, including the optional pod_ids query parameter and the structure of the APIPodStatusResponse.

```yaml
openapi: 3.1.0
info:
  title: PI API
  version: 0.1.0
servers:
  - url: https://api.primeintellect.ai
security: []
paths:
  /api/v1/pods/status:
    get:
      tags:
        - Pods
      summary: Get Pods Status
      operationId: get_pods_status_api_v1_pods_status_get
      parameters:
        - name: pod_ids
          in: query
          required: false
          schema:
            type: array
            items:
              type: string
            description: List of pod ids
            default: []
            title: Pod Ids
          description: List of pod ids
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: >-
                  #/components/schemas/GenericResponse_List_APIPodStatusResponse__
        '401':
          description: Authorization failed
        '422':
          description: Invalid request data
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
      security:
        - HTTPBearer: []
components:
  schemas:
    GenericResponse_List_APIPodStatusResponse__:
      properties:
        data:
          anyOf:
            - items:
                $ref: '#/components/schemas/APIPodStatusResponse'
              type: array
            - type: 'null'
          title: Data
          description: Response data
        status:
          anyOf:
            - type: string
            - type: 'null'
          title: Status
          description: Response status
      type: object
      title: GenericResponse[List[APIPodStatusResponse]]
    ErrorResponse:
      properties:
        errors:
          items:
            $ref: '#/components/schemas/ErrorDetail'
          type: array
          title: Errors
      type: object
      required:
        - errors
      title: ErrorResponse
    APIPodStatusResponse:
      properties:
        podId:
          type: string
          title: Podid
          description: Unique identifier for the pod.
        providerType:
          $ref: '#/components/schemas/ProviderTypeEnum'
          description: ID of the provider.
        status:
          $ref: '#/components/schemas/PodStatusEnum'
          description: Current status of the pod.
        sshConnection:
          anyOf:
            - type: string
            - items:
                anyOf:
                  - type: string
                  - type: 'null'
              type: array
            - type: 'null'
          title: Sshconnection
          description: >-
            SSH connection details for the pod, which may be a string or list of
            strings (if there are multiple nodes).
          examples:
            - root@135.23.125.123 -p 22
        ip:
          anyOf:
            - type: string
            - items:
                anyOf:
                  - type: string
                  - type: 'null'
              type: array
            - type: 'null'
          title: Ip
          description: >-
            IP address or list (if there are multiple nodes) of IP addresses
            associated with the pod.
          examples:
            - 135.23.125.123
        installationStatus:
          anyOf:
            - $ref: '#/components/schemas/InstallationStatusEnum'
            - type: 'null'
          description: Installation status of the pod, based on InstallationStatusEnum.
        installationFailure:
          anyOf:
            - type: string
            - type: 'null'
          title: Installationfailure
          description: >-
            Details about any installation failures that occurred, if
            applicable.
        installationProgress:
          anyOf:
            - type: integer
            - type: 'null'
          title: Installationprogress
          description: >-
            Percentage of the installation process that has been completed, if
            available.
      type: object
      required:
        - podId
        - providerType
        - status
      title: APIPodStatusResponse
    ErrorDetail:
      properties:
        param:
          type: string
          title: Param
        details:
          type: string
          title: Details
      type: object
      required:
        - param
        - details
      title: ErrorDetail
    ProviderTypeEnum:
      type: string
      enum:
        - runpod
        - fluidstack
        - lambdalabs
        - hyperstack
        - oblivus
        - cudocompute
        - scaleway
        - tensordock
        - datacrunch
        - latitude
        - crusoecloud
        - massedcompute
        - akash
        - primeintellect
```

--------------------------------

### OpenAPI Specification for Get Pods History

Source: https://docs.primeintellect.ai/api-reference/pods/get-pods-history

Defines the GET request structure, query parameters, and response schemas for retrieving pod history.

```yaml
openapi: 3.1.0
info:
  title: PI API
  version: 0.1.0
servers:
  - url: https://api.primeintellect.ai
security: []
paths:
  /api/v1/pods/history:
    get:
      tags:
        - Pods
      summary: Get Pods History
      operationId: get_pods_history_api_v1_pods_history_get
      parameters:
        - name: sort_by
          in: query
          required: false
          schema:
            enum:
              - terminatedAt
              - createdAt
            type: string
            default: terminatedAt
            title: Sort By
        - name: sort_order
          in: query
          required: false
          schema:
            enum:
              - asc
              - desc
            type: string
            default: desc
            title: Sort Order
        - name: offset
          in: query
          required: false
          schema:
            type: integer
            default: 0
            title: Offset
        - name: limit
          in: query
          required: false
          schema:
            type: integer
            default: 100
            title: Limit
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: >-
                  #/components/schemas/GenericPaginationResponse_List_HistoryObj__
        '401':
          description: Authorization failed
        '422':
          description: Invalid request data
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
      security:
        - HTTPBearer: []
components:
  schemas:
    GenericPaginationResponse_List_HistoryObj__:
      properties:
        total_count:
          type: integer
          title: Total Count
          description: Total number of items available in the dataset
          default: 0
        offset:
          type: integer
          minimum: 0
          title: Offset
          description: Number of items to skip before starting to collect the result set
          default: 0
        limit:
          type: integer
          minimum: 0
          title: Limit
          description: Maximum number of items to return
          default: 100
        data:
          items:
            $ref: '#/components/schemas/HistoryObj'
          type: array
          title: Data
        status:
          anyOf:
            - type: string
            - type: 'null'
          title: Status
          description: Response status
      type: object
      required:
        - data
      title: GenericPaginationResponse[List[HistoryObj]]
    ErrorResponse:
      properties:
        errors:
          items:
            $ref: '#/components/schemas/ErrorDetail'
          type: array
          title: Errors
      type: object
      required:
        - errors
      title: ErrorResponse
    HistoryObj:
      properties:
        id:
          type: string
          title: Id
          description: Unique identifier for the record.
        name:
          type: string
          title: Name
          description: Name assigned to the resource.
        providerType:
          $ref: '#/components/schemas/ProviderTypeEnum'
          description: ID of the provider.
        provisionedBy:
          anyOf:
            - type: string
            - type: 'null'
          title: Provisionedby
          description: Name of the user who provisioned the resource.
        type:
          $ref: '#/components/schemas/ResourceType'
          description: Type of the resource.
        createdAt:
          type: string
          format: date-time
          title: Createdat
          description: Timestamp when the resource was created.
        terminatedAt:
          anyOf:
            - type: string
              format: date-time
            - type: 'null'
          title: Terminatedat
          description: Timestamp when the resource was terminated.
        gpuName:
          type: string
          title: Gpuname
          description: Model of the GPU used in the resource.
        gpuCount:
          type: integer
          title: Gpucount
          description: Number of GPUs allocated to the resource.
          examples:
            - 1
        socket:
          anyOf:
            - type: string
            - type: 'null'
          title: Socket
          description: Type of socket used by the GPU, if applicable.
        priceHr:
          type: number
          title: Pricehr
          description: Price per hour for using the resource (average).
          examples:
            - 1.23
        userId:
          type: string
          title: Userid
```

--------------------------------

### Execute CUA Mode with Binary Upload

Source: https://docs.primeintellect.ai/guides/browser-environments

Run the CUA example using a custom server binary uploaded at runtime.

```bash
prime eval run browser-cua-example -m openai/gpt-4.1-mini -a '{"use_prebuilt_image": false}'
```

--------------------------------

### GET /api/v1/evaluations/

Source: https://docs.primeintellect.ai/api-reference/evals/list-evaluations

Get a list of evaluations owned by the authenticated user or their teams. By default, returns all evaluations the user has access to (personal + all teams).

```APIDOC
## GET /api/v1/evaluations/

### Description
Get a list of evaluations owned by the authenticated user or their teams. By default, returns all evaluations the user has access to (personal + all teams).

### Method
GET

### Endpoint
/api/v1/evaluations/

### Parameters
#### Query Parameters
- **team_id** (string) - Optional - Filter by specific team ID
- **environment_id** (string) - Optional - Filter by environment ID
- **environment_name** (string) - Optional - Filter by environment name
- **suite_id** (string) - Optional - Filter by suite ID
- **skip** (integer) - Optional - Number of items to skip (default: 0)
- **limit** (integer) - Optional - Maximum number of items to return (default: 50, max: 100)

### Response
#### Success Response (200)
- **evaluations** (array) - List of evaluation objects
- **total** (integer) - Total count of evaluations
- **skip** (integer) - Number of items skipped
- **limit** (integer) - Limit applied to the request
```

--------------------------------

### Start a frozen model server

Source: https://docs.primeintellect.ai/prime-rl/training

Launch a vLLM-based frozen model server for use with algorithms requiring a teacher or sampling source.

```bash
CUDA_VISIBLE_DEVICES=1 uv run inference \
  --vllm.model <frozen-model> --server.port 8001
```

--------------------------------

### OpenAPI Specification for Get Tunnel Status

Source: https://docs.primeintellect.ai/api-reference/tunnel/get-tunnel-status-endpoint

Defines the GET endpoint for retrieving tunnel status, including required parameters, response schemas, and security requirements.

```yaml
openapi: 3.1.0
info:
  title: PI API
  version: 0.1.0
servers:
  - url: https://api.primeintellect.ai
security: []
paths:
  /api/v1/tunnel/{tunnel_id}:
    get:
      tags:
        - Tunnel
      summary: Get Tunnel Status Endpoint
      description: Get status of a specific tunnel.
      operationId: get_tunnel_status_endpoint_api_v1_tunnel__tunnel_id__get
      parameters:
        - name: tunnel_id
          in: path
          required: true
          schema:
            type: string
            title: Tunnel Id
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TunnelStatusResponse'
        '401':
          description: Authorization failed
        '422':
          description: Invalid request data
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
      security:
        - HTTPBearer: []
components:
  schemas:
    TunnelStatusResponse:
      properties:
        tunnel_id:
          type: string
          title: Tunnel Id
          description: Unique tunnel identifier
        name:
          anyOf:
            - type: string
            - type: 'null'
          title: Name
          description: Friendly name if set
        status:
          $ref: '#/components/schemas/TunnelStatus'
          description: Current tunnel status
        hostname:
          type: string
          title: Hostname
          description: Tunnel hostname
        url:
          type: string
          title: Url
          description: Full HTTPS URL
        local_port:
          type: integer
          title: Local Port
          description: Local port forwarded by the tunnel
        labels:
          items:
            type: string
          type: array
          title: Labels
          description: Tunnel labels
        http_user:
          anyOf:
            - type: string
            - type: 'null'
          title: Http User
          description: HTTP basic auth username, if auth is enabled
        auth_enabled:
          type: boolean
          title: Auth Enabled
          description: Whether HTTP basic auth protects tunnel traffic
          default: false
        created_at:
          type: string
          format: date-time
          title: Created At
          description: Tunnel creation time
        connected_at:
          anyOf:
            - type: string
              format: date-time
            - type: 'null'
          title: Connected At
          description: When the client connected
        expires_at:
          type: string
          format: date-time
          title: Expires At
          description: Token expiration time
        terminated_at:
          anyOf:
            - type: string
              format: date-time
            - type: 'null'
          title: Terminated At
          description: When the tunnel was terminated
        user_id:
          type: string
          title: User Id
          description: Owner user ID
        team_id:
          anyOf:
            - type: string
            - type: 'null'
          title: Team Id
          description: Team ID if applicable
      type: object
      required:
        - tunnel_id
        - status
        - hostname
        - url
        - local_port
        - created_at
        - expires_at
        - user_id
      title: TunnelStatusResponse
      description: Response containing tunnel status information.
    ErrorResponse:
      properties:
        errors:
          items:
            $ref: '#/components/schemas/ErrorDetail'
          type: array
          title: Errors
      type: object
      required:
        - errors
      title: ErrorResponse
    TunnelStatus:
      type: string
      enum:
        - pending
        - connected
        - disconnected
        - expired
        - terminated
      title: TunnelStatus
    ErrorDetail:
      properties:
        param:
          type: string
          title: Param
        details:
          type: string
          title: Details
      type: object
      required:
        - param
        - details
      title: ErrorDetail
  securitySchemes:
    HTTPBearer:
      type: http
      scheme: bearer
```
Tokens:
Raw