Runware

https://runware.ai/docs
Admin
Runware is an AI media generation API platform offering image and audio generation, manipulation,...
Tokens:7,582,226
Snippets:37,982
Trust Score:8.8
Update:1 month ago
Show doc for...
Tokens:
Raw
### Oyster Hatchery Dusk Relight Guide CLI Example

Source: https://runware.ai/docs/models/controlnet-preprocess-seg/examples

Run the segmentation model using the Runware CLI tool.

```bash
runware run runware:controlnet-preprocess@seg \
  inputs.image=https://assets.runware.ai/assets/inputs/b14013c1-fdfa-4597-9538-502b7bc8049b.jpg
```

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

### Install Runware SDK

Source: https://runware.ai/docs/platform/javascript-legacy.md

Install the SDK using npm or yarn.

```bash
npm install @runware/sdk-js
```

```bash
yarn add @runware/sdk-js
```

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

### Install SDK with bun

Source: https://runware.ai/docs/platform/typescript.md

Install the Runware SDK using the bun package manager.

```bash
bun add @runware/sdk
```

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

### Install Runware SDK

Source: https://runware.ai/docs/platform/python-legacy.md

Install the package using pip.

```bash
pip install runware
```

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

### Install full skill catalog via OpenClaw

Source: https://runware.ai/docs/platform/skills.md

Install the entire skill catalog directly from the repository source.

```bash
openclaw skills install git:runware/runware-skills
```

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

### Install SDK with pnpm

Source: https://runware.ai/docs/platform/typescript.md

Install the Runware SDK using the pnpm package manager.

```bash
pnpm add @runware/sdk
```

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

### Install SDK with yarn

Source: https://runware.ai/docs/platform/typescript.md

Install the Runware SDK using the yarn package manager.

```bash
yarn add @runware/sdk
```

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

### Runware CLI Example

Source: https://runware.ai/docs/models/openai-clip-vit-l-14/examples

Example of using the Runware CLI to run a caption task.

```bash
runware run runware:151@1 \
  prompt="Describe this image accurately in one concise caption, identifying the main subject, warehouse environment, notable equipment and safety details, and the visible action. Do not infer hazards or events that are not clearly shown." \
  inputs.image=https://assets.runware.ai/assets/inputs/ecc21092-cadb-4da9-a4f3-3a6c07a406bd.jpg
```

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

### TypeScript SDK Museum Audio Guide

Source: https://runware.ai/docs/models/alibaba-qwen3-tts-1-7b-voicedesign/examples

Example of using the Runware TypeScript SDK to generate an audio guide for a museum planetarium.

```typescript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'alibaba:qwen@3-tts-1.7b-voicedesign',
  positivePrompt: 'A composed, knowledgeable female voice in her forties with a clear international English accent, gentle resonance, measured pacing, and soft breath control. The tone is welcoming and intellectually curious, like a senior museum guide leading visitors beneath a domed star theater. Subtle awe on cosmic phrases, crisp diction on scientific terms, never theatrical, never robotic. Warm, polished, intimate public-speaking quality with natural pauses and graceful sentence endings.',
  speech: {
    text: 'Welcome to the Hall of Ancient Skies. Above you, the ceiling recreates the winter constellations as they were charted over a thousand years ago. Notice how Orion stands slightly tilted compared with modern maps; this projection follows the perspective used by navigators who crossed open water with nothing but memory, geometry, and nerve. To your left is a bronze armillary sphere, its rings modeling the heavens as a set of interlocking paths. It is not only an instrument, but a theory made visible. As the lights dim, watch the red point near the horizon. That is Mars, rendered as it appeared during a famously bright opposition. For many cultures, such moments were more than astronomical events. They were signals, arguments, omens, and invitations to ask a daring question: if the sky changes, what else might change with it?',
    voice: 'design',
    language: 'English',
    speed: 0.94
  },
  settings: {
    maxNewTokens: 2048
  }
})
```

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

### Initialize and Run OpenAI GPT-5 Mini with Runware SDK

Source: https://runware.ai/docs/models/openai-gpt-5-mini/examples

Demonstrates how to create a client, connect to the Runware service, and execute a prompt using the openai:gpt@5-mini model with a specialized system prompt for grant writing.

```TypeScript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'openai:gpt@5-mini',
  settings: {
    systemPrompt: 'You are a senior nonprofit grant writer. Produce persuasive, funder-ready prose grounded strictly in the supplied facts. Never invent statistics, partnerships, commitments, or evaluation results. Preserve requested headings and limits, connect activities to measurable outcomes, and use clear language rather than inflated claims.',
    maxTokens: 5000,
    thinkingLevel: 'medium'
  },
  messages: [
    {
      role: 'user',
      content: 'I need help drafting a grant request for Harbor Skills Alliance, a workforce nonprofit in Baltimore. We are seeking support to expand our certified nursing assistant training program. What information do you need?'
    },
    {
      role: 'assistant',
      content: 'Please provide the funding request and grant period; community need; target population; program activities and timeline; recruitment and employer partners; projected outputs and outcomes; evaluation methods; organizational qualifications; total project budget and use of grant funds; and the funder\'s priorities, required headings, and word limit. I will avoid unsupported claims and mark any essential missing information.'
    },
    {
      role: 'user',

```

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

### ControlNet Configuration Example

Source: https://runware.ai/docs/models/alibaba-z-image-turbo/schema.json

Example of configuring a ControlNet model with a guide image and specific step constraints to influence image structure.

```json
"controlNet": [
  {
    "model": "<controlnet-model-air>",
    "guideImage": "c64351d5-4c59-42f7-95e1-eace013eddab",
    "weight": 0.7,
    "startStep": 0,
    "endStep": 20,
    "controlMode": "controlnet"
  }
]
```

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

### Runware CLI Example

Source: https://runware.ai/docs/models/bfl-flux-video-upscale/examples

Executes the video upscale task directly from the command line interface.

```bash
runware run bfl:flux@video-upscale \
  positivePrompt="Restore this compressed archival marquetry workshop footage with source-faithful detail suitable for a furniture museum master. Enhance the distinct grain and pores of walnut, maple, and dyed veneer; the crisp boundaries of the floral inlay pieces; fine saw teeth; tiny sawdust curls; scratched cutting-board fibers; tarnished brass fittings; paper pattern edges; knit cuff texture; and natural detail in the craftsperson’s hands. Improve edge definition and material separation while reducing block artifacts, ringing, chroma bleed, mosquito noise, and muddy shadow detail. Preserve the craftsperson’s identity, hand anatomy, tool geometry, original framing, coherent motion, warm tungsten illumination, restrained late-1980s analog color character, and original audio. Do not redesign the inlay, invent markings, alter the workshop, or add modern objects, text, graphics, excessive sharpening halos, plastic skin, or unstable frame-to-frame detail." \
  upscaleFactor=2 \
  settings.creativity=1 \
  inputs.video=https://assets.runware.ai/assets/inputs/d183482d-0803-4dd1-9b74-7124aeab91c8.mp4
```

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

### IP Adapters Configuration Example

Source: https://runware.ai/docs/models/bfl-flux-1-dev/schema.json

An example showing how to define multiple IP Adapters within the ipAdapters array, specifying the model, guide images, and weight for each.

```json
"ipAdapters": [
  {
    "model": "<ip-adapter-model-air>",
    "guideImages": ["c64351d5-4c59-42f7-95e1-eace013eddab"],
    "weight": 0.75
  },
  {
    "model": "<ip-adapter-model-air>",
    "guideImages": ["d7e8f9a0-2b5c-4e7f-a1d3-9c8b7a6e5d4f"],
    "weight": 0.5
  }
]
```

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

### PuLID Configuration Example

Source: https://runware.ai/docs/models/bfl-flux-1-dev/schema.json

Example configuration for the PuLID object, showing how to specify reference images, identity weight, CFG scale, and the starting step for identity influence.

```json
"puLID": {
  "images": ["59a2edc2-45e6-429f-be5f-7ded59b92046"],
  "idWeight": 1,
  "trueCFGScale": 1.5,
  "CFGStartStep": 3
}
```

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

### Initialize and run openai:gpt@5-mini with Runware SDK

Source: https://runware.ai/docs/models/openai-gpt-5-mini/examples

This TypeScript snippet demonstrates how to create a client using an API key from environment variables, connect to the service, and execute a prompt using the openai:gpt@5-mini model with a medium thinking level.

```typescript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'openai:gpt@5-mini',
  settings: {
    thinkingLevel: 'medium'
  },
  messages: [
    {
      role: 'user',
      content: 'You are advising the producers of TIDELINE, a completed 92-minute documentary about a Louisiana fishing community adapting to coastal erosion. Prepare a substantive distribution decision memo using only the facts below.\n\nPRODUCER OBJECTIVES, IN PRIORITY ORDER\n1. Protect the film’s credibility and community relationships.\n2. Secure enough near-term cash to satisfy the September 30 obligations described below.\n3. Achieve broad US audience reach within 12 months.\n4. Preserve educational and community-screening access at affordable prices.\n5. Retain meaningful approval or consultation over edits, key art, trailer, and publicity framing.\n6. Maximize long-term economics, without relying on speculative performance forecasts.\n\nFINANCIAL POSITION\n- A $420,000 bridge loan must be repaid by September 30. Assume the stated amount includes all interest and fees.\n- The production also owes $40,000 in deferred crew compensation on September 30.\n- The sales agent receives 10% of every license fee, minimum guarantee, or advance paid to the producer, deducted when each payment is received. For this exercise, do not apply the commission to later backend distributions.\n- Ignore taxes and the time value of money.\n- The producer currently has no unrestricted cash available for these obligations.\n\nOFFER A — NORTHLIGHT RELEASING\n- Territory: United States and Canada.\n- Rights: all media, including theatrical, television, TVOD, SVOD, AVOD, educational, non-theatrical, airlines, and community screenings.\n- Term: 12 years.\n- Minimum guarantee: $550,000, recoupable; 50% on signing and 50% on delivery acceptance.\n- Northlight says signing can occur by August 20, but delivery acceptance may take up to 60 days after delivery. '
```

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

### Initialize Runware Client and Execute Request

Source: https://runware.ai/docs/models/openai-gpt-5-mini/examples

Demonstrates how to authenticate using an environment variable and send a request to the openai:gpt@5-mini model.

```python
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "outputFormat": "JSON",
            "model": "openai:gpt@5-mini",
            "messages": [
                {
                    "role": "user",
```

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

### Museum Exhibition Audio Guide Script Generation

Source: https://runware.ai/docs/models/minimax-m2-7-highspeed/examples

A multi-turn conversation example where a user provides instructions and source material to generate a structured audio guide script.

```json
[
    {
        "role": "user",
        "content": "Write the complete audio guide for “After the Last Train,” a museum exhibition about photographer Elena Varga’s nighttime documentation of regional rail stations. I’ll provide the approved curatorial facts. The script will be recorded by a professional narrator, so it must be polished and ready for the studio."
    },
    {
        "role": "assistant",
        "content": "Please send the approved facts, required stops, target runtime, intended audience, interpretive tone, accessibility guidance, pronunciation notes, and any claims or language that must be avoided."
    },
    {
        "role": "user",
        "content": "Use only the source packet below. Do not invent quotations, dates, locations, motives, biographical details, or photographic techniques.\n\nDELIVERABLE\nCreate a six-stop English-language audio guide totaling 1,050–1,200 words. Write for curious adult visitors without specialist art knowledge. The experience should feel intimate, observant, and quietly hopeful—not academic, sentimental, eerie, or promotional.\n\nReturn the finished script only, organized as:\n- Exhibition Introduction\n- Stop 1: Platform Clock, 23:17\n- Stop 2: Waiting Room with Fern\n- Stop 3: Signal Box Window\n- Stop 4: Last Bus Connection\n- Stop 5: First Light, Empty Track\n- Closing\n\nFor every section, include an estimated runtime in brackets beneath the heading. Write natural narration in complete paragraphs. Do not include production notes, music cues, stage directions, footnotes, bullet points, or citations in the script.\n\nSOURCE PACKET\nExhibition title: After the Last Train\nArtist: Elena Varga (pronounced EH-leh-nah VAR-gah)\nVenue: Northbank Museum of Photography\n\nThe exhibition contains 42 color photographs made from 2018 through 2023 at small, still-operating railway stations across Hungary, Slovakia, Croatia, and eastern Austria. Varga traveled only on scheduled public transportation and photographed between the final evening arrival and the first morning departure. She did not stage scenes or ask station staff to alter lighting.\n\nVarga began the series after noticing that stations change function late at night: they remain pieces of public infrastructure, but briefly become places of maintenance, waiting, handover, and rest. Her stated interest is in “the hours when a shared place belongs to very few people.” This is the only direct quotation approved for use.\n\nAcross the series, recurring details include clocks, ticket windows, potted plants, vending machines, fluorescent fixtures, handwritten notices, benches, reflective safety clothing, cleaning equipment, and tracks disappearing beyond the lit platform. People are present in 11 of the 42 photographs. No individual is identified by name.\n\nVarga used a tripod and long exposures where station rules allowed. The photographs retain mixed light sources rather than correcting them to a uniform white. As a result, fluorescent light may appear green, older lamps amber, and early daylight blue. Long exposure sometimes softens a moving figure, but the architecture and stationary objects remain precise.\n\nThe works are presented in chronological order from late evening toward dawn rather than by country. Labels omit exact station names at the artist’s request. She wanted visitors to attend first to repeated civic features rather than treat the series as a catalog of destinations.\n\nOBJECT NOTES\nIntroduction area: A wall-sized route diagram connects all photographed stations but does not associate individual pictures with named locations.\n\nStop 1, Platform Clock, 23:17, 2018: A round analog clock occupies the upper center of the frame. Below it, two empty benches face opposite directions. Rainwater reflects a strip of amber platform light. This is the earliest photograph in the series.\n\nStop 2, Waiting Room with Fern, 2019: A large fern sits between a radiator and a closed ticket window. A handwritten notice is visible but not legible. Green fluorescent light fills the room, while an adjoining corridor appears amber. No people are visible.\n\nStop 3, Signal Box Window, 2020: Photographed from a public platform toward an illuminated signal box. One railway worker is visible in profile behind glass, wearing reflective clothing and looking down at paperwork. The long exposure softens one raised hand. Avoid speculating about the worker’s identity, thoughts, emotions, or exact task.\n\nStop 4, Last Bus Connection, 2021: Three passengers stand beneath a bus shelter outside a station. One checks a paper timetable, one holds a folded bicycle, and one sits beside two shopping bags. Their faces are not clearly visible. The station clock reads 00:42."
    }
]
```

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

### Woodcut Organ Restyle Detail Guide Response

Source: https://runware.ai/docs/models/controlnet-preprocess-tile/examples

Example JSON response from the API.

```json
{
  "taskType": "imageControlNetPreProcess",
  "taskUUID": "0f01190d-d5a2-449e-ab3f-9ea751cd0311",
  "guideImageURL": "https://im.runware.ai/image/os/a03d21/ws/3/ii/052868ab-29ea-4f44-8385-a2d539743b53.jpg",
  "cost": 0.0006,
  "guideImageUUID": "052868ab-29ea-4f44-8385-a2d539743b53",
  "inputImageUUID": "32634bd4-ebce-4d70-b6c2-54b1ee13b545"
}
```

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

### Generate a product explainer video with HeyGen Video Agent

Source: https://runware.ai/docs/models/heygen-video-agent/examples

This TypeScript example demonstrates how to initialize the Runware client and request a 15-second product explainer video using the heygen:video-agent@0 model.

```typescript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'heygen:video-agent@0',
  positivePrompt: 'Create a polished 15-second landscape product explainer for a reusable environmental DNA water-sampling kit used by river restoration teams. Use a calm female narrator and realistic documentary footage with precise macro product shots. Visual setting: a cool, overcast mountain river, slate-gray rocks, clear moving water, mossy banks, and a field scientist wearing practical charcoal waterproof gear and bright orange gloves. The compact sampler consists of a transparent syringe-style water chamber, a white sterile filter cartridge, an orange locking cap, and a small printed sample code. No logos or invented branding.\n\nStructure the video as a fast, coherent three-step demonstration. Open with a wide river establishing shot, then a close shot of the scientist filling the chamber with river water. Cut to an extreme macro view as water is pushed through the sterile filter, with a restrained animated flow line clarifying the direction. Show the cartridge twisting securely into its orange cap. Finish with the scientist recording the sample code on a waterproof field sheet, followed by a clean hero shot of the sealed cartridge beside the river.\n\nSpoken message, verbatim: “Collect biodiversity data in three steps. Draw river water through the sterile filter, twist the cartridge closed, then log its sample code. The preservative stabilizes environmental DNA on site, ready for transport to the lab.”\n\nPresentation style: premium natural-history documentary blended with crisp scientific instruction. Cool mineral-blue and gray palette with safety-orange accents, soft diffused daylight, shallow depth of field, subtle water ambience, minimal modern captions, and confident rhythmic editing. Add only three brief functional labels synchronized to the actions: “1 COLLECT”, “2 SEAL”, and “3 LOG”. Keep every product action physically credible, hands clearly visible, and the sampler consistent across shots. End on the sealed cartridge hero shot without a logo, slogan, call-to-action, app interface, or website.',
  width: 1280,
  height: 720,
  duration: 15
})
```

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

### Configure IP-Adapters

Source: https://runware.ai/docs/models/noobai.md

Example of defining multiple IP-Adapters with specific models, guide images, and weights.

```json
"ipAdapters": [
  {
    "model": "<ip-adapter-model-air>",
    "guideImages": ["c64351d5-4c59-42f7-95e1-eace013eddab"],
    "weight": 0.75
  },
  {
    "model": "<ip-adapter-model-air>",
    "guideImages": ["d7e8f9a0-2b5c-4e7f-a1d3-9c8b7a6e5d4f"],
    "weight": 0.5
  }
]
```

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

### CLI Command Example

Source: https://runware.ai/docs/models/bytedance-omnihuman-1-5/examples

Executes the model using the Runware CLI tool with specified provider settings and inputs.

```bash
runware run bytedance:5@2 \
  providerSettings.bytedance.fastMode=true \
  inputs.image=https://assets.runware.ai/assets/inputs/fc3d1850-0307-4ca9-9d27-772772d4460e.jpg \
  inputs.audio=https://assets.runware.ai/assets/inputs/a5bd1665-95c6-4be1-9783-7cad92608af1.mp3 \
  inputs.mask=https://assets.runware.ai/assets/inputs/0c621334-e6b2-4e10-886a-8deaea86fc17.png
```

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

### Initialize and run openai:gpt@5.4-mini with TypeScript

Source: https://runware.ai/docs/models/openai-gpt-5-4-mini/examples

Demonstrates how to create a client, connect to the Runware API, and submit a complex prompt for architectural decision-making.

```typescript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'openai:gpt@5.4-mini',
  settings: {
    thinkingLevel: 'medium'
  },
  messages: [
    {
      role: 'user',
      content: 'Act as a principal software architect advising Northstar Classroom, a B2B SaaS company that bills 8,000 school districts. Write a polished architecture decision memo for engineering, finance, security, and executive reviewers.\n\nCurrent state:\n- A 12-year-old Ruby on Rails monolith owns subscriptions, invoices, credits, payments, refunds, and revenue adjustments in one PostgreSQL database.\n- Peak volume is 55 transactions per second during annual renewals.\n- Billing jobs sometimes time out, but the system has had no confirmed ledger data loss.\n- Finance closes the books within five business days and requires an immutable audit trail.\n- Payment card data is tokenized by the payment processor; the company must preserve PCI scope boundaries.\n- Roughly 40 internal workflows and 17 downstream integrations read billing tables directly.\n- The engineering team has 14 backend developers and cannot pause roadmap work for more than one quarter.\n- Leadership wants the billing domain separated within 12 months, without a flag-day cutover.\n- Recovery objectives are RPO under five minutes and RTO under 30 minutes.\n\nEvaluate these three approaches:\n1. Rewrite the billing domain as event-sourced services and migrate all workflows at once.\n2. Extract a modular billing service with its own relational ledger, using change data capture during migration.\n3. Keep the monolith as system of record and add an API façade plus read replicas.\n\nRecommend one approach. Be candid about tradeoffs and do not treat fashionable architecture as inherently superior. Do not invent costs, compliance certifications, staffing, or performance results beyond the facts provided. Clearly label any assumptions.\n\nProduce 1,400–1,800 words in Markdown with these sections:\n- Decision\n- Executive Summary\n- Context and Non-Negotiable Constraints\n- Options Considered (include a compact comparison table covering delivery risk, auditability, reversibility, operational complexity, and 12-month feasibility)\n- Recommended Target Architecture\n- Migration Plan (four phases, each with entry criteria, exit criteria, and rollback trigger)\n- Data Integrity and Reconciliation Controls\n- Security and PCI Boundary Considerations\n- Observability and Service-Level Indicators\n- Principal Risks and Mitigations\n- Decisions Deferred\n- Approval Checklist\n\nThe target architecture must explicitly address idempotency, ordering, duplicate delivery, schema evolution, direct database consumers, backfills, dual-running reconciliation, disaster recovery, and ownership of monetary state. Define measurable indicators for ledger divergence, posting latency, failed transactions, reconciliation lag, and recovery readiness. End with a concise recommendation suitable for an executive approval meeting.'
    }
  ]
})
```

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

### JSON Response: Night Market Cooktop Scene Guide

Source: https://runware.ai/docs/models/controlnet-preprocess-shuffle/examples

Example response from the shuffle preprocessing task.

```json
{
  "taskType": "imageControlNetPreProcess",
  "taskUUID": "598a322b-b6d3-4682-804b-b4e12f733540",
  "guideImageURL": "https://im.runware.ai/image/os/a07dlim3/ws/3/ii/26555ddb-2367-432a-b617-d964876fec45.jpg",
  "cost": 0.0006,
  "guideImageUUID": "26555ddb-2367-432a-b617-d964876fec45",
  "inputImageUUID": "c8f09ec0-70fc-4df3-bac4-8ec6604f1914"
}
```

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

### Oyster Hatchery Dusk Relight Guide cURL Example

Source: https://runware.ai/docs/models/controlnet-preprocess-seg/examples

Direct API request to the Runware endpoint for segmentation.

```bash
curl https://api.runware.ai/v1 \
  -H "Authorization: Bearer $RUNWARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '[
    {
      "taskType": "controlNetPreprocess",
      "taskUUID": "fe99cb72-5ae5-4a57-b6d4-e870116fb5e4",
      "model": "runware:controlnet-preprocess@seg",
      "inputs": {
        "image": "https://assets.runware.ai/assets/inputs/b14013c1-fdfa-4597-9538-502b7bc8049b.jpg"
      }
    }
  ]'
```

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

### Initialize Runware Client and Execute Kimi K2.6

Source: https://runware.ai/docs/models/moonshotai-kimi-k2-6/examples

Demonstrates how to connect to the Runware SDK and run the Kimi K2.6 model with specific system prompts and image inputs for incident analysis.

```typescript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'moonshotai:kimi@k2.6',
  seed: 79285,
  settings: {
    systemPrompt: 'You are Kimi K2.6, an expert multimodal incident commander for complex infrastructure failures. Inspect visual evidence carefully, distinguish direct observations from inferences, and produce concise but actionable technical guidance.',
    temperature: 0.35,
    topP: 0.9,
    thinkingLevel: 'high',
    maxTokens: 8192
  },
  inputs: {
    images: [
      'https://assets.runware.ai/assets/inputs/2d5e8a55-51e2-4f3a-b1d4-dc81f78909bc.jpg'
    ]
  },
  messages: [
    {
      role: 'user',
      content: 'Analyze the attached image as if it is the only field photo available from a polar research facility outage. Provide: 1) a bullet list of concrete visual observations, 2) the three most likely failure chains with confidence levels, 3) immediate safety risks, 4) a step-by-step recovery plan for the next 60 minutes, 5) what evidence is missing, and 6) a final JSON object summarizing affected systems, severity, probable root cause, and recommended next actions. Do not invent unreadable text; mark uncertain details clearly.'
    }
  ]
})
```

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

### Geology Exhibition Panorama Expansion CLI Example

Source: https://runware.ai/docs/models/controlnet-preprocess-shuffle/examples

Uses the Runware CLI to run the shuffle model for the second example.

```bash
runware run runware:controlnet-preprocess@shuffle \
  inputs.image=https://assets.runware.ai/assets/inputs/df8a67b4-b192-476c-8004-2db0d2222fb1.jpg
```

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

### Configure Refiner Model

Source: https://runware.ai/docs/models/illustrious.md

Example configuration for the refiner object, specifying the model identifier and the step at which the refiner should start.

```json
"refiner": {
  "model": "<refiner-model-air>",
  "startStep": 30
}
```

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

### Initialize and run Kimi K2.6 with TypeScript

Source: https://runware.ai/docs/models/moonshotai-kimi-k2-6/examples

Demonstrates how to create a Runware client using an API key and execute a model request with specific system prompt settings and conversation history.

```typescript
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'moonshotai:kimi@k2.6',
  settings: {
    systemPrompt: 'You are Kimi K2.6, a helpful, careful, conversational assistant. Ask clarifying questions when needed, reason through tradeoffs, and give practical guidance in a calm tone. Keep the exchange natural and supportive.',
    temperature: 0.45,
    topP: 0.9,
    thinkingLevel: 'medium',
    maxTokens: 8192
  },
  messages: [
    {
      role: 'user',
      content: 'Hi, I need help making sense of a kitchen remodel. I have a small 1980s kitchen, a budget of about $18,000, and I keep getting conflicting advice from contractors. Can you help me figure out what questions I should ask and how to prioritize the work?'
    },
    {
      role: 'assistant',
      content: 'Absolutely. I can help you organize the project into priorities, budget categories, contractor questions, and decision points. To make the advice useful, tell me a bit more: are you changing the layout, replacing cabinets, moving plumbing or electrical, or mainly updating surfaces and appliances?'
    },
    {
      role: 'user',
      content: 'I want to keep the layout mostly the same, but I might remove a half wall between the kitchen and dining area. Cabinets are worn out, counters are laminate, flooring is old vinyl, and the lighting is terrible. Appliances are functional but mismatched. What should I do first, and how do I avoid spending the whole budget too early?'
    }
  ]
})
```

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

### Espresso Machine Colorway Edge Guide CLI

Source: https://runware.ai/docs/models/controlnet-preprocess-softedge/examples

Example using the Runware CLI to process an espresso machine image.

```bash
runware run runware:controlnet-preprocess@softedge \
  inputs.image=https://assets.runware.ai/assets/inputs/a3cf5f9b-2d12-4da2-bb3e-7485a1d8b7f6.jpg
```

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

### Initialize and Run Minimax M2-7 Model

Source: https://runware.ai/docs/models/minimax-m2-7/examples

Demonstrates how to authenticate with the Runware client and execute a prompt using the minimax:m2.7@0 model with specific operational strategy parameters.

```python
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "minimax:m2.7@0",
            "seed": 66209,
            "settings": {
                "maxTokens": 24000,
                "temperature": 0.62,
                "topP": 0.9,
                "systemPrompt": "You are a senior product operations and launch strategist. In multi-turn conversations, preserve prior context, revise rather than restart, and produce executive-ready outputs that are concise, structured, and operationally grounded. Prioritize decision clarity, dependencies, risk tradeoffs, and measurable go/no-go criteria."
            },
            "messages": [
                {
                    "role": "user",
```

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

### Oyster Hatchery Dusk Relight Guide Python Example

Source: https://runware.ai/docs/models/controlnet-preprocess-seg/examples

Uses the Runware Python client to process an image for segmentation.

```python
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "runware:controlnet-preprocess@seg",
            "inputs": {
                "image": "https://assets.runware.ai/assets/inputs/b14013c1-fdfa-4597-9538-502b7bc8049b.jpg"
            }
        })


asyncio.run(main())
```
Tokens:
Raw