OpenAI Compatibility

deAPI’s OpenAI-compatible API gives you access to image generation, speech synthesis, transcription, video generation, and embeddings — at lower cost, on decentralized GPU infrastructure. Switch from OpenAI by changing two parameters. Your existing code stays unchanged. This is the same approach used by Groq, Together AI, Fireworks AI, and other leading inference providers.

Quick start

Python
Node.js / TypeScript
cURL

from openai import OpenAI

client = OpenAI(
    api_key="dpn-sk-your-token-here",
    base_url="https://oai.deapi.ai/v1"
)

# Generate an image — same call as OpenAI
response = client.images.generate(
    model="Flux1schnell",
    prompt="A futuristic city at sunset, cinematic lighting",
    size="1024x1024",
    n=1
)
print(response.data[0].url)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "dpn-sk-your-token-here",
  baseURL: "https://oai.deapi.ai/v1",
});

// Generate an image — same call as OpenAI
const response = await client.images.generate({
  model: "Flux1schnell",
  prompt: "A futuristic city at sunset, cinematic lighting",
  size: "1024x1024",
  n: 1,
});
console.log(response.data[0].url);

curl -X POST "https://oai.deapi.ai/v1/images/generations" \
  -H "Authorization: Bearer dpn-sk-your-token-here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Flux1schnell",
    "prompt": "A futuristic city at sunset, cinematic lighting",
    "size": "1024x1024",
    "n": 1
  }'

Get your API key at app.deapi.ai/dashboard. New accounts receive a $5 bonus — no credit card required.

What changes vs OpenAI

Parameter	OpenAI	deAPI
`base_url` / `baseURL`	`https://api.openai.com/v1`	`https://oai.deapi.ai/v1`
`api_key` / `apiKey`	`sk-...`	Your deAPI key (starts with `dpn-sk-`)
`model`	e.g. `dall-e-3`	deAPI model ID (e.g. `Flux1schnell`)

Your full API key looks like dpn-sk-<id>|<secret> — the dpn-sk- prefix is required by the gateway, and applies only here (the native API at https://api.deapi.ai takes the raw key). You can find it in your Dashboard → Settings → API Keys. Everything else — request format, response schema, error envelope — follows the OpenAI specification.

Available models

deAPI runs open-source models, not OpenAI’s proprietary ones. Model IDs are native slugs (e.g. Flux1schnell, Kokoro, WhisperLargeV3) — not OpenAI model names. This is by design. Use GET /v1/models to fetch the current list:

curl "https://oai.deapi.ai/v1/models" \
  -H "Authorization: Bearer dpn-sk-your-token-here"

Response follows the OpenAI format:

{
  "object": "list",
  "data": [
    { "id": "Flux1schnell", "object": "model", "created": 1700000000, "owned_by": "deapi" },
    { "id": "Kokoro", "object": "model", "created": 1700000000, "owned_by": "deapi" }
  ]
}

Models are added and updated regularly. For the full, always-current list with capabilities, limits, and defaults, see the Model Selection endpoint and the Models guide.

Supported endpoints

Endpoint	Status	Description
`GET /v1/models`	✅	List available models
`POST /v1/images/generations`	✅	Text-to-image
`POST /v1/images/edits`	✅	Image editing (img2img)
`POST /v1/audio/speech`	✅	Text-to-speech
`POST /v1/audio/transcriptions`	✅	Audio & video transcription
`POST /v1/embeddings`	✅	Text embeddings
`POST /v1/videos`	✅	Video generation
`POST /v1/chat/completions`	❌	Not in scope — deAPI does not serve LLMs
`POST /v1/images/edits` with `mask`	❌	Inpainting not supported — returns 400
`POST /v1/files`	🔜	Coming soon — files are sent inline (multipart) for now

Migration examples

Image generation (DALL-E → deAPI)

Before (OpenAI):

from openai import OpenAI

client = OpenAI(api_key="sk-...")

response = client.images.generate(
    model="dall-e-3",
    prompt="A cozy cabin in the woods",
    size="1024x1024",
    n=1
)

After (deAPI) — change api_key, base_url, and model:

from openai import OpenAI

client = OpenAI(
    api_key="dpn-sk-your-token-here",       # ← changed
    base_url="https://oai.deapi.ai/v1"   # ← changed
)

response = client.images.generate(
    model="Flux1schnell",                 # ← deAPI model ID
    prompt="A cozy cabin in the woods",
    size="1024x1024",
    n=1
)

Text-to-speech (OpenAI TTS → deAPI)

Before (OpenAI):

audio = client.audio.speech.create(
    model="tts-1",
    voice="alloy",
    input="Hello world"
)

After (deAPI):

from openai import OpenAI

client = OpenAI(
    api_key="dpn-sk-your-token-here",
    base_url="https://oai.deapi.ai/v1"
)

audio = client.audio.speech.create(
    model="Kokoro",
    voice="alloy",   # OpenAI voice aliases supported: alloy, echo, fable, onyx, nova, shimmer
    input="Hello world"
)

Supported output format: mp3 only. Any other response_format is rejected with 400.

Kokoro supports the same six OpenAI voice aliases (alloy, echo, fable, onyx, nova, shimmer). Voice language is determined by the voice prefix, not the input text: af_/am_ → US English, bf_/bm_ → British English. See the TTS endpoint docs for the full voice list.

Transcription (Whisper → deAPI)

Before (OpenAI):

with open("audio.mp3", "rb") as f:
    transcript = client.audio.transcriptions.create(
        model="whisper-1",
        file=f
    )
print(transcript.text)

After (deAPI):

from openai import OpenAI

client = OpenAI(
    api_key="dpn-sk-your-token-here",
    base_url="https://oai.deapi.ai/v1"
)

with open("audio.mp3", "rb") as f:
    transcript = client.audio.transcriptions.create(
        model="WhisperLargeV3",
        file=f
    )
print(transcript.text)

Supported response_format values: "json" (default), "text", "verbose_json". The maximum audio file size is 20 MB, subject to a global 75 MB request body cap. See Limits & Quotas → File Uploads for the full breakdown.

Embeddings

Before (OpenAI):

response = client.embeddings.create(
    model="text-embedding-3-small",
    input="The quick brown fox"
)

After (deAPI):

from openai import OpenAI

client = OpenAI(
    api_key="dpn-sk-your-token-here",
    base_url="https://oai.deapi.ai/v1"
)

response = client.embeddings.create(
    model="Bge_M3_FP16",
    input="The quick brown fox"
)

# Vector dimension: 1024
print(len(response.data[0].embedding))  # → 1024

Supports single string and array input. Supports encoding_format: "base64".

Framework integrations

Because deAPI uses the OpenAI API format, it works with any framework that accepts a base_url / api_base parameter.

LangChain

from langchain_openai import OpenAIEmbeddings

embeddings = OpenAIEmbeddings(
    openai_api_key="dpn-sk-your-token-here",
    openai_api_base="https://oai.deapi.ai/v1",
    model="Bge_M3_FP16"
)

LlamaIndex

from llama_index.embeddings.openai import OpenAIEmbedding

embed_model = OpenAIEmbedding(
    api_key="dpn-sk-your-token-here",
    api_base="https://oai.deapi.ai/v1",
    model="Bge_M3_FP16"
)

Vercel AI SDK

import { createOpenAI } from "@ai-sdk/openai";

const deapi = createOpenAI({
  apiKey: "dpn-sk-your-token-here",
  baseURL: "https://oai.deapi.ai/v1",
});

Environment variables

If your codebase already uses the standard OpenAI environment variables, override them at the process level — zero code changes required:

export OPENAI_API_KEY="dpn-sk-your-token-here"
export OPENAI_BASE_URL="https://oai.deapi.ai/v1"

For multi-provider setups (e.g. OpenAI for chat, deAPI for images and audio), instantiate separate clients:

import os
from openai import OpenAI

openai_client = OpenAI(api_key=os.environ["OPENAI_KEY"])

deapi_client = OpenAI(
    api_key=os.environ["DEAPI_KEY"],
    base_url="https://oai.deapi.ai/v1",
)

# openai_client → GPT models
# deapi_client  → images, audio, video, embeddings

Known differences

Feature	OpenAI	deAPI
Chat completions	✅	❌ Out of scope
Inpainting (`mask` in `/v1/images/edits`)	✅	❌ Returns 400
Model IDs	`dall-e-3`, `tts-1`, `whisper-1`	Native slugs (e.g. `Flux1schnell`, `Kokoro`)
Image `size` values	OpenAI fixed set	Model-specific — check Models
Max `n` (images)	10	4
Audio file size limit	25 MB	20 MB (per-file) — total request body capped at 75 MB
Embedding dimensions	1536 (ada-002)	Model-specific (e.g. 1024 for `Bge_M3_FP16`)
`style: "vivid"`	Controls image style	Accepted but ignored
Video generation	❌	✅ `POST /v1/videos`
`/v1/files`	✅	🔜 Coming soon

For the full list of models, supported parameters, and limits, see the Model Selection endpoint.

Next steps

Models

Discover all available models and their capabilities

Quickstart

Get your API key and make your first request

API v2 Reference

Full native API reference with all endpoints

Pricing

Pay-as-you-go rates per task and model

Get started

Execution Modes & Integrations

Other

OpenAI Compatibility

Quick start

What changes vs OpenAI

Available models

Supported endpoints

Migration examples

Image generation (DALL-E → deAPI)

Text-to-speech (OpenAI TTS → deAPI)

Transcription (Whisper → deAPI)

Embeddings

Framework integrations

LangChain

LlamaIndex

Vercel AI SDK

Environment variables

Known differences

Next steps

Models

Quickstart

API v2 Reference

Pricing

​Quick start

​What changes vs OpenAI

​Available models

​Supported endpoints

​Migration examples

​Image generation (DALL-E → deAPI)

​Text-to-speech (OpenAI TTS → deAPI)

​Transcription (Whisper → deAPI)

​Embeddings

​Framework integrations

​LangChain

​LlamaIndex

​Vercel AI SDK

​Environment variables

​Known differences

​Next steps

Models

Quickstart

API v2 Reference

Pricing

Quick start

What changes vs OpenAI

Available models

Supported endpoints

Migration examples

Image generation (DALL-E → deAPI)

Text-to-speech (OpenAI TTS → deAPI)

Transcription (Whisper → deAPI)

Embeddings

Framework integrations

LangChain

LlamaIndex

Vercel AI SDK

Environment variables

Known differences

Next steps