Connect AI agents to the Runware MCP server

Vendor MCP

The Runware MCP connector routes your AI agent's tool calls to Runware's own MCP server through Scalekit. Each user signs in to Runware once, and Scalekit stores and refreshes their tokens, so your agent never handles credentials. It comes with 12 tools.

Tools: 12
What they doRead · write · destructive: 9 · 2 · 19 read2 write1 destructive
Users sign in with: OAuth
OAuth app: Scalekit's or your own

Setup

Install the SDK
- Node.js
- Python
Terminal window
npm install @scalekit-sdk/node dotenv
Terminal window
pip install scalekit-sdk-python python-dotenv
Set your credentials
Add your Scalekit credentials to your .env file. Find values in app.scalekit.com > Developers > API Credentials.
.env
```
SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
SCALEKIT_CLIENT_ID=<your-client-id>
SCALEKIT_CLIENT_SECRET=<your-client-secret>
```
Create the Runware MCP connection
In AgentKit > Connections, create a Runware MCP connection. The name you give it is the connection_name your code passes. See Configure connections.
Scalekit credentials are available for Runware MCP server, so you don't need to register an OAuth app.

Authorize a user and make your first call

Node.js
Python

1
import { ScalekitClient } from '@scalekit-sdk/node'
2
import 'dotenv/config'
3
import { createInterface } from 'node:readline/promises'
4

5
const scalekit = new ScalekitClient(
6
  process.env.SCALEKIT_ENVIRONMENT_URL,
7
  process.env.SCALEKIT_CLIENT_ID,
8
  process.env.SCALEKIT_CLIENT_SECRET,
9
)
10
const actions = scalekit.actions
11

12
const connector = 'runwaremcp'
13
const identifier = 'user_123'
14

15
// Generate an authorization link for the user
16
const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
17
console.log('Authorize Runware MCP:', link)
18
const rl = createInterface({ input: process.stdin, output: process.stdout })
19
await rl.question('Press Enter after authorizing...')
20
rl.close()
21

22
// Make your first call
23
const result = await actions.executeTool({
24
  connector,
25
  identifier,
26
  toolName: 'runwaremcp_list_capabilities',
27
  toolInput: {},
28
})
29
console.log(result)

npx tsx quickstart.mts

1
import os
2
from scalekit import ScalekitClient
3
from dotenv import load_dotenv
4
load_dotenv()
5

6
scalekit_client = ScalekitClient(
7
    env_url=os.getenv("SCALEKIT_ENVIRONMENT_URL"),
8
    client_id=os.getenv("SCALEKIT_CLIENT_ID"),
9
    client_secret=os.getenv("SCALEKIT_CLIENT_SECRET"),
10
)
11
actions = scalekit_client.actions
12

13
connection_name = "runwaremcp"
14
identifier = "user_123"
15

16
# Generate an authorization link for the user
17
link_response = actions.get_authorization_link(
18
    connection_name=connection_name,
19
    identifier=identifier,
20
)
21
print("Authorize Runware MCP:", link_response.link)
22
input("Press Enter after authorizing...")
23

24
# Make your first call
25
result = actions.execute_tool(
26
    tool_input={},
27
    tool_name="runwaremcp_list_capabilities",
28
    connection_name=connection_name,
29
    identifier=identifier,
30
)
31
print(result)

python quickstart.py

Each user signs in once. See Authorize a user for the full flow and statuses.

Tools

Pass the exact name to execute_tool

Try in Playground Request a tool

runwaremcp_accountRetrieve Runware account information including balance and usage.Read-only
Account
Retrieve Runware account information including balance and usage.
Inputs
operationstringrequired
Which account operation to perform. Currently only "getDetails" is supported.one of getDetails
includeCostboolean
Include cost information
runwaremcp_get_task_detailsRetrieve the original request and response for a previously executed task.Read-only
GetTaskDetails
Retrieve the original request and response for a previously executed task. Useful for recovering results or auditing past generations.
Inputs
taskUUIDstringrequired
UUID of the task to retrieve
runwaremcp_list_capabilitiesList every model capability Runware supports, with their human-readable labels.Read-only
List Capabilities
List every model capability Runware supports, with their human-readable labels. Use this to discover the taxonomy (e.g. "io:text-to-image", "op:upscale") before filtering list_models by capability, or to answer "what can Runware do?".
Inputs
This tool takes no inputs.
runwaremcp_list_modelsList Runware's official, curated model integrations.Read-only
List Models
List Runware's official, curated model integrations. Returns each model's name, AIR identifier, headline, capabilities, and pricing. Call this FIRST whenever the user names or asks about a model that could be first-party (e.g. "FLUX 2 dev", "SDXL", "Veo 3", "Gemma", "Wan 2.5", "Z-Image"), and also for open-ended "what models are available?" questions. Match the user's named model against the returned names. Only fall through to model_search if no curated entry matches.
Inputs
capabilitystring
Optional capability filter (e.g. "io:text-to-image", "op:upscale"). Use list_capabilities first to discover valid ids.
categorystring
Optional output-modality filterone of imagevideoaudiotext3d
creatorstring
Optional creator id filter (e.g. "google", "alibaba")
searchstring
Optional free-text search across name, AIR, creator, capabilities
runwaremcp_model_detailsGet the full curated metadata for a single Runware model by AIR identifier — name, headline, description, capabilities, pricing, and creator.Read-only
Get Model Details
Get the full curated metadata for a single Runware model by AIR identifier — name, headline, description, capabilities, pricing, and creator. Use this when the user wants more depth on a model already surfaced by list_models, or to confirm an AIR matches what the user named.
Inputs
airstringrequired
Model AIR identifier (e.g. "runware:400@1")
runwaremcp_model_examplesGet sample input/output examples for a curated Runware model.Read-only
Get Model Examples
Get sample input/output examples for a curated Runware model. Useful when the user wants to see what a model produces, or to crib a working request shape before constructing a run() call.
Inputs
airstringrequired
Model AIR identifier
capabilitystring
Optional capability filter (e.g. "io:text-to-image")
runwaremcp_model_pricingGet pricing details for a curated Runware model — overview text plus example configurations with prices (e.g.Read-only
Get Model Pricing
Get pricing details for a curated Runware model — overview text plus example configurations with prices (e.g. "1024×1024 = $0.0032"). Use this when the user asks how much a specific model will cost.
Inputs
airstringrequired
Model AIR identifier
runwaremcp_model_schemaGet the parameter schema for a specific model.Read-only
Get Model Schema
Get the parameter schema for a specific model. Returns the JSON Schema describing all accepted parameters, their types, defaults, and constraints. ALWAYS call this before calling run() with a model you haven't used before, so you know what parameters to pass.
Inputs
modelstringrequired
Model AIR identifier (e.g., "runware:400@1", "google:3@3", "google:gemma@4-31b")
runwaremcp_model_searchSearch Runware's Civitai mirror and community-uploaded models — third-party fine-tunes, user uploads, style LoRAs, custom checkpoints.Read-only
ModelSearch
Search Runware's Civitai mirror and community-uploaded models — third-party fine-tunes, user uploads, style LoRAs, custom checkpoints. ONLY use this AFTER list_models has been checked and the user's named model is not in the curated catalog, OR when the user explicitly asks for a Civitai or community model. Do NOT use this for first-party models like FLUX, SDXL, Veo, Imagen, Gemma, Wan, Z-Image — those live in list_models.
Inputs
searchstringrequired
Search query (name, description, or AIR ID)
architecturestring
Filter by architecture (e.g. "flux-1-dev", "sdxl")
categorystring
Filter by model categoryone of checkpointloralycorisvaeembeddings
tagsarray
Filter by tags
typestring
Filter checkpoints by type (only when category=checkpoint)one of baseinpaintingrefiner
runwaremcp_model_uploadUpload a custom AI model to Runware (checkpoint, LoRA, VAE, embeddings, etc.).Write
ModelUpload
Upload a custom AI model to Runware (checkpoint, LoRA, VAE, embeddings, etc.). Returns the AIR identifier once the upload completes.
Inputs
architecturestringrequired
Model architecture (e.g. "flux-1-dev", "sdxl")
categorystringrequired
Model categoryone of checkpointloralycorisvaeembeddings
downloadURLstringrequired
URL where the model weights can be downloaded from
formatstringrequired
Weight file formatone of safetensors
namestringrequired
Display name for the model
versionstringrequired
Model version
airstring
Optional AIR identifier (format: provider:model@version)
runwaremcp_runRun an AI inference task on Runware.Write
Run
Run an AI inference task on Runware. Supports image generation, video generation, audio generation, 3D generation, upscaling, background removal, captioning, and more. Pass a model AIR identifier and task-specific parameters. Example: { "model": "runware:400@1", "positivePrompt": "a cat", "width": 1024, "height": 1024 }
Inputs
modelstringrequired
Model AIR identifier (e.g., "runware:400@1" for image, "google:3@3" for video, "google:gemma@4-31b" for LLM)
runwaremcp_media_storageStore or delete media (images, video, audio, 3D models) in your Runware account.Destructive
Media storage
Store or delete media (images, video, audio, 3D models) in your Runware account. Upload returns a media UUID you can reuse as input (seedImage, referenceImages, etc.); delete removes previously stored media by its media UUID.
Inputs
mediastringrequired
For upload: URL, data URI, or base64 of the media. For delete: the media UUID to remove
operationstringrequired
Whether to upload new media or delete stored mediaone of uploaddelete

Setup

Install the SDK

Set your credentials

Create the Runware MCP connection

Authorize a user and make your first call

Tools

Call it

Returns

Account

Inputs

GetTaskDetails

Inputs

List Capabilities

Inputs

List Models

Inputs

Get Model Details

Inputs

Get Model Examples

Inputs

Get Model Pricing

Inputs

Get Model Schema

Inputs

ModelSearch

Inputs

ModelUpload

Inputs

Run

Inputs

Media storage

Inputs

Related: other AI connectors