Skill: Deapi Github

Owner: zrewolwerowanykaloryfer

Summary: The cheapest AI media API on the market. Transcribe YouTube videos, generate images with Flux and Z-Image models, convert text to speech in 54+ voices across 8 languages, extract text with OCR, create videos, remove backgrounds, upscale images, apply style transfer - all through one unified API. Free $5 credit on signup - enough for hundreds of hours of transcription or thousands of generated images. Fraction of the cost of any alternative.

Tags: latest:1.0.9

Version history:

v1.0.9 | 2026-02-13T18:24:49.020Z | user

Reverted to 1.0.6

v1.0.8 | 2026-02-13T18:12:42.696Z | user

Reverted 1.0.7 changes (untested LTX-2 update)

v1.0.7 | 2026-02-13T18:11:01.769Z | user

Added LTX-2 19B video model

v1.0.6 | 2026-02-11T13:57:37.783Z | user

Reverted to placeholder paths per Testie review

v1.0.5 | 2026-02-11T13:21:26.631Z | user

Security fix: explicit /tmp/deapi_*.png paths instead of placeholders to pass virus scan

v1.0.4 | 2026-02-11T13:08:41.979Z | user

Fixed metadata format for ClawdHub requiredEnv detection (clawdbot.requires.env)

v1.0.3 | 2026-02-11T12:25:01.912Z | user

Fixed source URL to correct repository, updated author

v1.0.2 | 2026-02-11T12:00:22.316Z | user

Added homepage, source, author, license to metadata. Added privacy considerations (result_url accessibility, data handling). Added provenance links.

v1.0.1 | 2026-02-11T11:57:34.104Z | user

Security improvements: replaced /tmp/ paths with placeholders, added security note explaining the skill documents legitimate API

v1.0.0 | 2026-02-11T11:46:37.770Z | user

Initial release - transcription (YouTube/audio/video), image generation (Flux, Z-Image), TTS (54+ voices, 8 languages), OCR, video generation, background removal, upscale (4x), style transfer, embeddings. 10-20x cheaper than alternatives.

Archive index:

Archive v1.0.9: 3 files, 6181 bytes

Files: README.md (2780b), SKILL.md (12905b), _meta.json (124b)

File v1.0.9:SKILL.md

---

name: deAPI AI Media Suite description: The cheapest AI media API on the market. Transcribe YouTube videos, generate images with Flux and Z-Image models, convert text to speech in 54+ voices across 8 languages, extract text with OCR, create videos, remove backgrounds, upscale images, apply style transfer - all through one unified API. Free $5 credit on signup - enough for hundreds of hours of transcription or thousands of generated images. Fraction of the cost of any alternative. homepage: https://deapi.ai source: https://github.com/zrewolwerowanykaloryfer/deapi-clawdbot-skill author: zrewolwerowanykaloryfer license: MIT requiredEnv:

• DEAPI_API_KEY metadata: {"clawdbot":{"requires":{"env":["DEAPI_API_KEY"]}}} tags:
• media
• transcription
• image-generation
• tts
• ocr
• video
• audio
• embeddings

---

deAPI Media Generation

AI-powered media tools via decentralized GPU network. Get your API key at deapi.ai (free $5 credit on signup).

Setup

export DEAPI_API_KEY=your_api_key_here

Available Functions

| Function | Use when user wants to... | |----------|---------------------------| | Transcribe | Transcribe YouTube, Twitch, Kick, X videos, or audio files | | Generate Image | Generate images from text descriptions (Flux models) | | Generate Audio | Convert text to speech (TTS, 54+ voices, 8 languages) | | Generate Video | Create video from text or animate images | | OCR | Extract text from images | | Remove Background | Remove background from images | | Upscale | Upscale image resolution (2x/4x) | | Transform Image | Apply style transfer to images (multi-image support) | | Embeddings | Generate text embeddings for semantic search | | Check Balance | Check account balance |

---

Async Pattern (Important!)

All deAPI requests are asynchronous. Follow this pattern for every operation:

1. Submit Request

curl -s -X POST "https://api.deapi.ai/api/v1/client/{endpoint}" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{...}'

Response contains request_id.

2. Poll Status (loop every 10 seconds)

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

3. Handle Status

• processing → wait 10s, poll again
• done → fetch result from result_url
• failed → report error to user

Common Error Handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check DEAPI_API_KEY | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once |

---

Transcription (YouTube, Audio, Video)

Use when: user wants to transcribe video from YouTube, X, Twitch, Kick or audio files.

Endpoints:

• Video (YouTube, mp4, webm): vid2txt
• Audio (mp3, wav, m4a, flac, ogg): aud2txt

Request (video):

curl -s -X POST "https://api.deapi.ai/api/v1/client/vid2txt" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"video_url": "{VIDEO_URL}", "include_ts": true, "model": "WhisperLargeV3"}'

Request (audio):

curl -s -X POST "https://api.deapi.ai/api/v1/client/aud2txt" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"audio_url": "{AUDIO_URL}", "include_ts": true, "model": "WhisperLargeV3"}'

After polling: Present transcription with timestamps in readable format.

---

Image Generation (Flux)

Use when: user wants to generate images from text descriptions.

Endpoint: txt2img

Models: | Model | API Name | Steps | Max Size | Notes | |-------|----------|-------|----------|-------| | Klein (default) | Flux_2_Klein_4B_BF16 | 4 (fixed) | 1536px | Fastest, recommended | | Flux | Flux1schnell | 4-10 | 2048px | Higher resolution | | Turbo | ZImageTurbo_INT8 | 4-10 | 1024px | Fastest inference |

Request:

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2img" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "{PROMPT}",
    "model": "Flux_2_Klein_4B_BF16",
    "width": 1024,
    "height": 1024,
    "steps": 4,
    "seed": {RANDOM_0_TO_999999}
  }'

Note: Klein model does NOT support guidance parameter - omit it.

---

Text-to-Speech (54+ Voices)

Use when: user wants to convert text to speech.

Endpoint: txt2audio

Popular Voices: | Voice ID | Language | Description | |----------|----------|-------------| | af_bella | American EN | Warm, friendly (best quality) | | af_heart | American EN | Expressive, emotional | | am_adam | American EN | Deep, authoritative | | bf_emma | British EN | Elegant (best British) | | jf_alpha | Japanese | Natural Japanese female | | zf_xiaobei | Chinese | Mandarin female | | ef_dora | Spanish | Spanish female | | ff_siwis | French | French female (best quality) |

Voice format: {lang}{gender}_{name} (e.g., af_bella = American Female Bella)

Request:

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2audio" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "{TEXT}",
    "voice": "af_bella",
    "model": "Kokoro",
    "lang": "en-us",
    "speed": 1.0,
    "format": "mp3",
    "sample_rate": 24000
  }'

Parameters:

• lang: en-us, en-gb, ja, zh, es, fr, hi, it, pt-br
• speed: 0.5-2.0
• format: mp3/wav/flac/ogg
• sample_rate: 22050/24000/44100/48000

---

Video Generation

Use when: user wants to generate video from text or animate an image.

Endpoints:

• Text-to-Video: txt2video (multipart/form-data)
• Image-to-Video: img2video (multipart/form-data)

⚠️ IMPORTANT: Model-specific constraints for Ltxv_13B_0_9_8_Distilled_FP8:

• guidance: MUST be 0 (max value!)
• steps: MUST be 1 (max value!)
• fps: minimum 30

Request (text-to-video):

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2video" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "prompt={PROMPT}" \
  -F "model=Ltxv_13B_0_9_8_Distilled_FP8" \
  -F "width=512" \
  -F "height=512" \
  -F "guidance=0" \
  -F "steps=1" \
  -F "frames=120" \
  -F "fps=30" \
  -F "seed={RANDOM_0_TO_999999}"

Parameters: | Parameter | Required | Constraints | Description | |-----------|----------|-------------|-------------| | prompt | Yes | - | Video description | | model | Yes | - | Ltxv_13B_0_9_8_Distilled_FP8 | | width | Yes | 256-768 | Video width (e.g., 512) | | height | Yes | 256-768 | Video height (e.g., 512) | | guidance | Yes | max 0 | Must be 0 for this model | | steps | Yes | max 1 | Must be 1 for this model | | frames | Yes | 30-300 | Number of frames | | fps | Yes | min 30 | Frames per second | | seed | Yes | 0-999999 | Random seed |

Request (image-to-video):

# Download image first if URL provided
curl -s -o {LOCAL_IMAGE_PATH} "{IMAGE_URL}"

curl -s -X POST "https://api.deapi.ai/api/v1/client/img2video" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "first_frame_image=@{LOCAL_IMAGE_PATH}" \
  -F "prompt=gentle movement, cinematic" \
  -F "model=Ltxv_13B_0_9_8_Distilled_FP8" \
  -F "width=512" \
  -F "height=512" \
  -F "guidance=0" \
  -F "steps=1" \
  -F "frames=120" \
  -F "fps=30" \
  -F "seed={RANDOM_0_TO_999999}"

Note: Video generation can take 1-3 minutes.

---

OCR (Image to Text)

Use when: user wants to extract text from an image.

Endpoint: img2txt (multipart/form-data)

Request:

# Download image first if URL provided
curl -s -o {LOCAL_IMAGE_PATH} "{IMAGE_URL}"

# Send OCR request
curl -s -X POST "https://api.deapi.ai/api/v1/client/img2txt" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{LOCAL_IMAGE_PATH}" \
  -F "model=Nanonets_Ocr_S_F16"

---

Background Removal

Use when: user wants to remove background from an image.

Endpoint: img-rmbg (multipart/form-data)

Request:

# Download image first if URL provided
curl -s -o {LOCAL_IMAGE_PATH} "{IMAGE_URL}"

# Send remove-bg request
curl -s -X POST "https://api.deapi.ai/api/v1/client/img-rmbg" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{LOCAL_IMAGE_PATH}" \
  -F "model=Ben2"

Result: PNG with transparent background.

---

Image Upscale (2x/4x)

Use when: user wants to upscale/enhance image resolution.

Endpoint: img-upscale (multipart/form-data)

Models: | Scale | Model | |-------|-------| | 2x | RealESRGAN_x2 | | 4x | RealESRGAN_x4 |

Request:

# Download image first if URL provided
curl -s -o {LOCAL_IMAGE_PATH} "{IMAGE_URL}"

# Send upscale request
curl -s -X POST "https://api.deapi.ai/api/v1/client/img-upscale" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{LOCAL_IMAGE_PATH}" \
  -F "model=RealESRGAN_x4"

---

Image Transformation (Style Transfer)

Use when: user wants to transform image style, combine images, or apply AI modifications.

Endpoint: img2img (multipart/form-data)

Models: | Model | API Name | Max Images | Guidance | Steps | Notes | |-------|----------|------------|----------|-------|-------| | Klein (default) | Flux_2_Klein_4B_BF16 | 3 | N/A (ignore) | 4 (fixed) | Faster, multi-image | | Qwen | QwenImageEdit_Plus_NF4 | 1 | 7.5 | 10-50 (default 20) | More control |

Request (Klein, supports up to 3 images):

# Download images first
curl -s -o {LOCAL_IMAGE_1} "{IMAGE_URL_1}"
curl -s -o {LOCAL_IMAGE_2} "{IMAGE_URL_2}"  # optional

# Send transform request (Klein - no guidance)
curl -s -X POST "https://api.deapi.ai/api/v1/client/img2img" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{LOCAL_IMAGE_1}" \
  -F "image=@{LOCAL_IMAGE_2}" \
  -F "prompt={STYLE_PROMPT}" \
  -F "model=Flux_2_Klein_4B_BF16" \
  -F "steps=4" \
  -F "seed={RANDOM_0_TO_999999}"

Request (Qwen, higher quality single image):

# Download image first
curl -s -o {LOCAL_IMAGE_1} "{IMAGE_URL}"

# Send transform request (Qwen - with guidance)
curl -s -X POST "https://api.deapi.ai/api/v1/client/img2img" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{LOCAL_IMAGE_1}" \
  -F "prompt={STYLE_PROMPT}" \
  -F "model=QwenImageEdit_Plus_NF4" \
  -F "guidance=7.5" \
  -F "steps=20" \
  -F "seed={RANDOM_0_TO_999999}"

Example prompts: "convert to watercolor painting", "anime style", "cyberpunk neon aesthetic"

---

Text Embeddings

Use when: user needs embeddings for semantic search, clustering, or RAG.

Endpoint: txt2embedding

Request:

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2embedding" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"input": "{TEXT}", "model": "Bge_M3_FP16"}'

Result: 1024-dimensional vector (BGE-M3, multilingual)

---

Check Balance

Use when: user wants to check remaining credits.

Request:

curl -s "https://api.deapi.ai/api/v1/client/balance" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Response: { "data": { "balance": 4.25 } }

---

Pricing (Approximate)

| Operation | Cost | |-----------|------| | Transcription | ~$0.02/hour | | Image Generation | ~$0.002/image | | TTS | ~$0.001/1000 chars | | Video Generation | ~$0.05/video | | OCR | ~$0.001/image | | Remove BG | ~$0.001/image | | Upscale | ~$0.002/image | | Embeddings | ~$0.0001/1000 tokens |

Free $5 credit on signup at deapi.ai.

---

Converted from deapi-ai/claude-code-skills for Clawdbot/OpenClaw.

---

Security & Privacy Note

This skill provides documentation for the deAPI.ai REST API, a legitimate decentralized AI media service.

Security:

• All curl commands are examples showing how to call the API
• Requests go to api.deapi.ai (official deAPI endpoint)
• Local file paths (e.g., {LOCAL_IMAGE_PATH}) are placeholders - use any suitable temporary location
• The skill itself does not execute code or download binaries
• API key is required and must be set by user via DEAPI_API_KEY environment variable

Privacy considerations:

• Media URLs you submit (YouTube links, images) are sent to deapi.ai for processing
• Generated results are returned via result_url which may be temporarily accessible via direct link
• Results are stored on deAPI's infrastructure - review their privacy policy for retention details
• Do not process sensitive/confidential media without understanding data handling

Provenance:

• Service provider: deapi.ai
• Original skill source: github.com/deapi-ai/claude-code-skills
• API documentation: docs.deapi.ai

File v1.0.9:README.md

deAPI AI Media Suite

The cheapest AI media API on the market - now as a Clawdbot/OpenClaw skill.

Transcribe YouTube videos, generate images, convert text to speech, extract text with OCR, create videos, remove backgrounds, and more - all through one unified API at a fraction of the cost.

🚀 Features

| Feature | Description | |---------|-------------| | 🎬 Transcription | YouTube, Twitch, Kick, X, audio/video files (Whisper) | | 🖼️ Image Generation | Flux and Z-Image models | | 🗣️ Text-to-Speech | 54+ voices across 8 languages | | 📝 OCR | Extract text from images | | 🎥 Video Generation | Text-to-video and image-to-video | | ✂️ Background Removal | Remove backgrounds from images | | 🔍 Upscale | 2x/4x image upscaling (RealESRGAN) | | 🎨 Style Transfer | Transform images with AI | | 🧮 Embeddings | Text embeddings for semantic search |

📦 Installation

clawdhub install deapi

⚙️ Setup

1. Get your API key at deapi.ai (free $5 credit on signup)
2. Set environment variable (or simply provide the API key to your agent when needed):

   export DEAPI_API_KEY=your_api_key_here
   

💰 Pricing

• Transcription: ~$0.02/hour
• Image Generation: ~$0.002/image
• TTS: ~$0.001/1000 chars
• Video Generation: ~$0.05/video

Free $5 credit = hundreds of hours of transcription or thousands of images!

📖 Usage

Once installed, your Clawdbot agent automatically knows how to use deAPI. Just ask:

• "Transcribe this YouTube video: [URL]"
• "Generate an image of a robot in cyberpunk style"
• "Convert this text to speech with a British accent"
• "Extract text from this screenshot"
• "Remove the background from this photo"
• "Upscale this image 4x"

🔒 Security & Privacy

• All requests go to api.deapi.ai (official endpoint)
• Media URLs you submit are sent to deAPI for processing
• Results via result_url may be temporarily accessible
• Review deAPI's privacy policy for data handling details

🙏 Credits

• API Provider: deAPI.ai
• Original Claude Code skill: deapi-ai/claude-code-skills
• Clawdbot conversion: @zrewolwerowanykaloryfer

📄 License

MIT License - see LICENSE for details.

---

Links: ClawHub | deAPI.ai | API Docs

File v1.0.9:_meta.json

{ "ownerId": "kn7d2s78ex44zmw98bdv67vsw180y6e0", "slug": "deapi", "version": "1.0.9", "publishedAt": 1771007089020 }

Archive v1.0.8: 13 files, 15839 bytes

Files: commands/deapi-balance.md (591b), commands/deapi-setup.md (3671b), commands/embed.md (2180b), commands/generate-audio.md (3397b), commands/generate-image.md (2687b), commands/generate-video.md (2346b), commands/ocr.md (2005b), commands/remove-bg.md (1972b), commands/transcribe.md (1900b), commands/transform-image.md (3468b), commands/upscale.md (2136b), SKILL.md (1830b), _meta.json (124b)

File v1.0.8:SKILL.md

---

name: deapi description: AI media generation via deAPI. Transcribe YouTube/audio/video, generate images from text, text-to-speech, OCR, remove backgrounds, upscale images, create videos, generate embeddings. 10-20x cheaper than OpenAI/Replicate. metadata: clawdbot: config: requiredEnv: - DEAPI_API_KEY tags:

• media
• video
• image-generation
• transcription
• tts
• ocr

---

deAPI Media Generation

AI-powered media tools via decentralized GPU network.

Available Commands

| Command | Use when user wants to... | |---------|---------------------------| | /transcribe | Transcribe YouTube, Twitch, Kick, X videos, or audio files | | /generate-image | Generate images from text descriptions | | /generate-audio | Convert text to speech (TTS) | | /generate-video | Create video from text or animate images | | /ocr | Extract text from images (OCR) | | /remove-bg | Remove background from images | | /upscale | Upscale image resolution (2x/4x) | | /transform-image | Apply style transfer to images | | /embed | Generate text embeddings for semantic search | | /deapi-setup | Configure result delivery (webhooks/websockets) for server apps | | /deapi-balance | Check account balance and remaining credits |

Quick Examples

/transcribe https://youtube.com/watch?v=...
/generate-image a sunset over mountains
/generate-audio "Hello world" --voice am_adam

Setup

Requires DEAPI_API_KEY environment variable:

export DEAPI_API_KEY=your_key

Get your API key at deapi.ai (free $5 credit).

API Pattern

All deAPI requests are async:

1. Submit job → get request_id
2. Poll status every 10s
3. When done → fetch from result_url

For detailed API parameters, see docs/api-reference.md.

File v1.0.8:_meta.json

{ "ownerId": "kn7d2s78ex44zmw98bdv67vsw180y6e0", "slug": "deapi", "version": "1.0.8", "publishedAt": 1771006362696 }

File v1.0.8:commands/deapi-balance.md

---

name: deapi-balance description: Check deAPI account balance

deAPI Account Balance

Step 1: Check balance

curl -s "https://api.deapi.ai/api/v1/client/balance" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Step 2: Present result

Response: { "data": { "balance": 4.25 } }

Display: deAPI balance: {balance} credits

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 403 Forbidden | API key may be revoked, get a new one at deapi.ai | | Network error | Check internet connection |

File v1.0.8:commands/deapi-setup.md

---

name: deapi-setup description: Configure result delivery method for deAPI jobs (polling, webhooks, websockets)

Result Delivery Setup

Configure how your application receives job results.

Step 1: Determine context

Ask user about their use case:

| Context | Recommended method | |---------|-------------------| | CLI tool, local script | Polling (default) | | Backend server, serverless | Webhooks | | Real-time UI, SPA | WebSockets |

Step 2: Configure based on method

Polling (default)

No configuration needed. Use /request-status/{id} endpoint. Already built into all deAPI commands.

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Poll every 10 seconds until status = "done".

Webhooks

Add webhook_url to any request:

{
  "prompt": "...",
  "webhook_url": "https://your-server.com/webhooks/deapi"
}

Events received:

• job.completed - contains result_url
• job.failed - contains error_code

Security: Verify signature in X-DeAPI-Signature header (HMAC-SHA256).

Full docs: https://docs.deapi.ai/execution-modes-and-integrations/webhooks

WebSockets

Connect via Pusher protocol for real-time updates:

| Setting | Value | |---------|-------| | Host | soketi.deapi.ai:443 | | Channel | private-client.{client_id} | | Event | request.status.updated |

Includes live preview images during generation.

Full docs: https://docs.deapi.ai/execution-modes-and-integrations/websockets

Step 3: Generate boilerplate

Based on user's choice, generate starter code for their stack.

Webhook handler (Node.js/Express):

app.post('/webhooks/deapi', (req, res) => {
  const signature = req.headers['x-deapi-signature'];
  // Verify signature with your webhook secret

  const { event, request_id, result_url, error } = req.body;

  if (event === 'job.completed') {
    // Process result from result_url
  } else if (event === 'job.failed') {
    // Handle error
  }

  res.status(200).send('OK');
});

Webhook handler (Python/FastAPI):

from fastapi import FastAPI, Request, Header

app = FastAPI()

@app.post("/webhooks/deapi")
async def deapi_webhook(
    request: Request,
    x_deapi_signature: str = Header(None)
):
    # Verify signature with your webhook secret
    body = await request.json()

    if body["event"] == "job.completed":
        result_url = body["result_url"]
        # Process result
    elif body["event"] == "job.failed":
        error = body["error"]
        # Handle error

    return {"status": "ok"}

WebSocket client (JavaScript):

import Pusher from 'pusher-js';

const pusher = new Pusher('your-app-key', {
  wsHost: 'soketi.deapi.ai',
  wsPort: 443,
  forceTLS: true,
  disableStats: true,
  enabledTransports: ['ws', 'wss']
});

const channel = pusher.subscribe(`private-client.${clientId}`);

channel.bind('request.status.updated', (data) => {
  if (data.status === 'done') {
    // Fetch result from data.result_url
  }
});

Error handling

| Error | Action | |-------|--------| | Webhook timeout | Retry with exponential backoff (1s, 2s, 4s) | | Invalid signature | Verify webhook secret matches dashboard config | | WebSocket disconnect | Reconnect with backoff, resubscribe to channel | | 401 on polling | Check DEAPI_API_KEY is valid | | Missing result_url | Job may still be processing, continue polling |

When to suggest this skill

Proactively suggest /deapi-setup when user mentions:

• Building a web app with deAPI
• Server-side integration
• Real-time UI updates
• Production deployment

File v1.0.8:commands/embed.md

---

name: embed description: Generate text embeddings for semantic search and similarity argument-hint: <text>

Text Embeddings via deAPI

Generate embeddings for: $ARGUMENTS

Step 1: Validate input

Verify $ARGUMENTS contains text to embed:

• Text should not be empty
• Maximum recommended length: 8192 tokens
• For longer texts, consider chunking

Step 2: Send request

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2embedding" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": "$ARGUMENTS",
    "model": "Bge_M3_FP16"
  }'

Model info: | Model | Dimensions | Best for | |-------|------------|----------| | Bge_M3_FP16 | 1024 | High accuracy, semantic search, multilingual |

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get embedding vector from response
2. Show vector dimensions and sample values
3. Offer to save or use the embedding

Output format:

{
  "embedding": [0.123, -0.456, 0.789, ...],
  "dimensions": 1024,
  "model": "Bge_M3_FP16"
}

Step 5: Offer follow-up

Ask user:

• "Would you like to embed more text for comparison?"
• "Should I calculate similarity between embeddings?"
• "Would you like to save these embeddings to a file?"

Use cases

• Semantic search: Find similar documents
• Clustering: Group related content
• RAG: Retrieval-augmented generation
• Recommendations: Content similarity

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Empty text | Ask user to provide text to embed | | Text too long | Suggest chunking into smaller segments |

File v1.0.8:commands/generate-audio.md

---

name: generate-audio description: Convert text to speech using 54+ AI voices in 8 languages argument-hint: <text> [--voice af_bella]

Text-to-Speech via deAPI

Generate audio from text: $ARGUMENTS

Step 1: Parse arguments

Extract from $ARGUMENTS:

• text: The text to convert to speech (required)
• --voice: Voice ID (default: af_bella)

54+ voices in 8 languages. Voice ID format: {lang}{gender}_{name}

• Lang: a=American, b=British, j=Japanese, z=Chinese, e=Spanish, f=French, h=Hindi, i=Italian, p=Portuguese
• Gender: f=female, m=male

Popular voices: | Voice ID | Language | Description | |----------|----------|-------------| | af_bella | American EN | Warm, friendly (best quality) | | af_heart | American EN | Expressive, emotional | | af_nicole | American EN | Professional | | af_sarah | American EN | Casual | | am_adam | American EN | Deep, authoritative | | am_michael | American EN | Energetic | | bf_emma | British EN | Elegant (best British) | | bf_isabella | British EN | Refined | | bm_george | British EN | Sophisticated | | jf_alpha | Japanese | Natural Japanese female | | zf_xiaobei | Chinese | Mandarin female | | ef_dora | Spanish | Spanish female | | ff_siwis | French | French female (best quality) | | hf_alpha | Hindi | Hindi female | | if_sara | Italian | Italian female | | pf_dora | Portuguese | Brazilian Portuguese female |

Full list: Kokoro VOICES.md

Step 2: Send request

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2audio" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "{text}",
    "voice": "{voice}",
    "model": "Kokoro",
    "lang": "{lang_code}",
    "speed": 1.0,
    "format": "mp3",
    "sample_rate": 24000
  }'

Parameters: | Parameter | Default | Description | |-----------|---------|-------------| | lang | auto from voice | Language code: en-us, en-gb, ja, zh, es, fr, hi, it, pt-br | | speed | 1.0 | Speech speed (0.5-2.0) | | format | mp3 | Output format: mp3, wav, flac, ogg | | sample_rate | 24000 | Sample rate: 22050, 24000, 44100, 48000 |

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get audio URL from result_url
2. Provide download link to user
3. Show audio duration if available

Step 5: Offer follow-up

Ask user:

• "Would you like to try a different voice?"
• "Should I generate this in another language?"
• "Would you like to adjust the text?"

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Empty text | Ask user to provide text to convert | | Invalid voice | Show available voices, ask user to choose | | Text too long | Suggest splitting into smaller segments |

File v1.0.8:commands/generate-image.md

---

name: generate-image description: Generate images from text prompts using Flux Klein, Flux Schnell, or ZImageTurbo models argument-hint: <prompt> [--model klein|flux|turbo] [--size 512|768|1024]

Image Generation via deAPI

Generate image from prompt: $ARGUMENTS

Step 1: Parse arguments

Extract from $ARGUMENTS:

• prompt: The text description (required)
• --model: klein (default, recommended), flux (higher max resolution), or turbo (fastest)
• --size: 512, 768, or 1024 (default: 1024). Klein supports up to 1536.

Step 2: Send request

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2img" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "{prompt}",
    "model": "{model_name}",
    "width": {size},
    "height": {size},
    "steps": {steps},
    "seed": {random_seed}
  }'

Model mapping: | User flag | API model name | Steps | Max size | Info | |-----------|----------------|-------|----------|------| | klein (default) | Flux_2_Klein_4B_BF16 | 4 (fixed) | 1536 | Recommended, txt2img + img2img, multi-image | | flux | Flux1schnell | 4-10 (default: 4) | 2048 | Higher max resolution | | turbo | ZImageTurbo_INT8 | 4-10 (default: 4) | 1024 | Fastest |

Klein model limits:

• Resolution: 256-1536px (step: 16)
• Steps: 4 (fixed)
• Guidance: not supported (parameter ignored)

Important: Generate a random seed (0-999999) for each request.

Note: Klein does not support guidance parameter - omit it or it will be ignored.

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get image URL from result_url
2. Display/download the generated image
3. Show the image to the user

Step 5: Offer follow-up

Ask user:

• "Would you like variations with a modified prompt?"
• "Should I upscale this image?"
• "Would you like to transform this with additional images?" (Klein supports up to 3 input images in img2img)

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Empty prompt | Ask user to provide a description | | NSFW rejected | Inform user, suggest alternative prompt |

File v1.0.8:commands/generate-video.md

---

name: generate-video description: Generate videos from text prompts or animate images argument-hint: <prompt-or-image-url> [--mode text|image]

Video Generation via deAPI

Generate video: $ARGUMENTS

Step 1: Determine mode and parse arguments

Auto-detect mode:

• If $ARGUMENTS starts with http → Image-to-Video (img2video)
• Otherwise → Text-to-Video (txt2video)

Override with --mode:

• --mode text: Force text-to-video
• --mode image: Force image-to-video

Step 2: Send request

For Text-to-Video:

curl -s -X POST "https://api.deapi.ai/api/v1/client/txt2video" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "{prompt}",
    "duration": 4,
    "fps": 24,
    "model": "Ltxv_13B_0_9_8_Distilled_FP8"
  }'

For Image-to-Video (animation):

curl -s -X POST "https://api.deapi.ai/api/v1/client/img2video" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "image_url": "{image_url}",
    "motion_prompt": "gentle movement, cinematic",
    "duration": 4,
    "fps": 24
  }'

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again (video generation can take 1-3 minutes)
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get video URL from result_url
2. Provide download link to user
3. Show video details (duration, resolution)

Step 5: Offer follow-up

Ask user:

• "Would you like to generate a longer version?"
• "Should I try a different motion style?"
• "Would you like to generate more variations?"

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Empty prompt | Ask user to provide a description | | Invalid image URL | Verify URL is accessible and points to an image | | Timeout (>5min) | Inform user, video generation is resource-intensive |

File v1.0.8:commands/ocr.md

---

name: ocr description: Extract text from images using AI-powered OCR argument-hint: <image-url>

OCR (Image to Text) via deAPI

Extract text from image: $ARGUMENTS

Step 1: Validate input

Verify $ARGUMENTS is a valid image file path or URL:

• Supported formats: .png, .jpg, .jpeg, .webp, .gif, .bmp
• If URL provided, download the file first

Step 2: Send request

Note: This endpoint requires multipart/form-data with file upload.

curl -s -X POST "https://api.deapi.ai/api/v1/client/img2txt" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{local_file_path}" \
  -F "model=Nanonets_Ocr_S_F16"

If user provides a URL, first download the image:

curl -s -o /tmp/ocr_image.png "{image_url}"

Then use /tmp/ocr_image.png as the file path.

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get extracted text from response or result_url
2. Present text in clean, formatted manner
3. Preserve original structure (paragraphs, lists) where possible

Step 5: Offer follow-up

Ask user:

• "Would you like me to summarize this text?"
• "Should I translate this to another language?"
• "Would you like to extract text from another image?"

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Invalid URL | Ask user to verify the image URL | | No text found | Inform user the image may not contain readable text | | Image too large | Suggest resizing or cropping the image |

File v1.0.8:commands/remove-bg.md

---

name: remove-bg description: Remove background from images using AI argument-hint: <image-url>

Background Removal via deAPI

Remove background from image: $ARGUMENTS

Step 1: Validate input

Verify $ARGUMENTS is a valid image file path or URL:

• Supported formats: .png, .jpg, .jpeg, .webp
• If URL provided, download the file first
• Best results with clear subject/foreground

Step 2: Send request

Note: This endpoint requires multipart/form-data with file upload.

curl -s -X POST "https://api.deapi.ai/api/v1/client/img-rmbg" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{local_file_path}" \
  -F "model=Ben2"

If user provides a URL, first download the image:

curl -s -o /tmp/rmbg_image.png "{image_url}"

Then use /tmp/rmbg_image.png as the file path.

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get processed image URL from result_url
2. Result is PNG with transparent background
3. Provide download link to user

Step 5: Offer follow-up

Ask user:

• "Would you like to add a new background color?"
• "Should I process another image?"
• "Would you like to resize the result?"

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Invalid URL | Ask user to verify the image URL | | No subject detected | Image may lack clear foreground; suggest different image | | Image too large | Suggest resizing before processing |

File v1.0.8:commands/transcribe.md

---

name: transcribe description: Transcribes video/audio from URLs (YouTube, X, mp3, mp4, wav) using Whisper AI argument-hint: <url>

Transcription via deAPI

Transcribe media from URL: $ARGUMENTS

Step 1: Determine endpoint

| URL type | Endpoint | |----------|----------| | .mp3, .wav, .m4a, .flac, .ogg | aud2txt | | YouTube, X, .mp4, .webm, other video | vid2txt |

Step 2: Send request

For video:

curl -s -X POST "https://api.deapi.ai/api/v1/client/vid2txt" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"video_url": "$ARGUMENTS", "include_ts": true, "model": "WhisperLargeV3"}'

For audio:

curl -s -X POST "https://api.deapi.ai/api/v1/client/aud2txt" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"audio_url": "$ARGUMENTS", "include_ts": true, "model": "WhisperLargeV3"}'

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done", fetch content from result_url and present transcription in readable format with timestamps.

Step 5: Offer follow-up

Ask user: "Would you like me to summarize the main points?"

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Invalid URL | Ask user to verify the URL format | | Timeout (>5min) | Inform user, suggest trying shorter content |

File v1.0.8:commands/transform-image.md

---

name: transform-image description: Transform images with style transfer and AI modifications argument-hint: <image-url> <style-prompt> [--model klein|qwen]

Image Transformation via deAPI

Transform image: $ARGUMENTS

Step 1: Parse arguments

Extract from $ARGUMENTS:

• image_url: URL(s) to source image(s) - Klein supports up to 3 images (required)
• style_prompt: Description of desired transformation (required)
• --model: klein (default, faster, multi-image) or qwen (higher steps, more control)

Example style prompts:

• "convert to watercolor painting"
• "make it look like a vintage photograph"
• "transform into anime style"
• "add cyberpunk neon aesthetic"
• "combine these images into one scene" (multi-image with Klein)

Step 2: Send request

Note: This endpoint requires multipart/form-data with file upload.

If user provides URL(s), first download the image(s):

curl -s -o /tmp/transform_image1.png "{image_url_1}"
curl -s -o /tmp/transform_image2.png "{image_url_2}"  # optional
curl -s -o /tmp/transform_image3.png "{image_url_3}"  # optional

Then send the transformation request:

curl -s -X POST "https://api.deapi.ai/api/v1/client/img2img" \
  -H "Authorization: Bearer $DEAPI_API_KEY" \
  -F "image=@{local_file_path_1}" \
  -F "image=@{local_file_path_2}" \
  -F "image=@{local_file_path_3}" \
  -F "prompt={style_prompt}" \
  -F "model={model_name}" \
  -F "guidance={guidance}" \
  -F "steps={steps}" \
  -F "seed={random_seed}"

Model mapping: | User flag | API model name | Steps | Guidance | Max images | Info | |-----------|----------------|-------|----------|------------|------| | klein (default) | Flux_2_Klein_4B_BF16 | 4 (fixed) | ignored | 3 | Faster, multi-image support | | qwen | QwenImageEdit_Plus_NF4 | 10-50 (default: 20) | 7.5 | 1 | More control, higher fidelity |

Klein model limits:

• Resolution: 256-1536px (step: 16)
• Steps: 4 (fixed)
• Guidance: not supported
• Max input images: 3

Parameters: | Parameter | Klein | Qwen | Description | |-----------|-------|------|-------------| | guidance | ignored | 7.5 | Prompt adherence (1-20) | | steps | 4 (fixed) | 20 | Inference steps | | seed | random | random | For reproducibility (0-999999) |

Step 3: Poll status (feedback loop)

Extract request_id from response, then poll every 10 seconds:

curl -s "https://api.deapi.ai/api/v1/client/request-status/{request_id}" \
  -H "Authorization: Bearer $DEAPI_API_KEY"

Status handling:

• processing → wait 10s, poll again
• done → proceed to Step 4
• failed → report error message to user, STOP

Step 4: Fetch and present result

When status = "done":

1. Get transformed image URL from result_url
2. Display before/after comparison if possible
3. Provide download link to user

Step 5: Offer follow-up

Ask user:

• "Would you like to try a different style?"
• "Should I add more input images?" (Klein only, up to 3)
• "Would you like to upscale the result?"

Error handling

| Error | Action | |-------|--------| | 401 Unauthorized | Check if $DEAPI_API_KEY is set correctly | | 429 Rate Limited | Wait 60s and retry | | 500 Server Error | Wait 30s and retry once | | Invalid URL | Ask user to verify the image URL | | Missing prompt | Ask user to describe desired transformation | | Too many images | Klein max 3, Qwen max 1 | | NSFW rejected | Inform user, suggest alternative style |