OpenAI API Relay

POST https://www.myai168.com/en/api/openai/v1/responses

Responses API (new chat interface with streaming support).

curl https://www.myai168.com/en/api/openai/v1/responses \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-sol","input":"Hello"}'

import requests

r = requests.post(
    "https://www.myai168.com/en/api/openai/v1/responses",
    headers={"Authorization": "Bearer YOUR_DEV_KEY"},
    json={"model": "gpt-5.6-sol", "input": "Hello"},
)
print(r.json())

POST https://www.myai168.com/en/api/openai/v1/chat/completions

Chat Completions API (legacy chat interface, OpenAI SDK compatible).

curl https://www.myai168.com/en/api/openai/v1/chat/completions \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-sol","messages":[{"role":"user","content":"Hello"}]}'

import requests

r = requests.post(
    "https://www.myai168.com/en/api/openai/v1/chat/completions",
    headers={"Authorization": "Bearer YOUR_DEV_KEY"},
    json={"model": "gpt-5.6-sol", "messages": [{"role": "user", "content": "Hello"}]},
)
print(r.json())

POST https://www.myai168.com/en/api/openai/v1/audio/speech

Text-to-speech (TTS). Returns audio binary.

curl https://www.myai168.com/en/api/openai/v1/audio/speech \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"tts-1","input":"Hello","voice":"nova"}' \
  --output speech.mp3

import requests

r = requests.post(
    "https://www.myai168.com/en/api/openai/v1/audio/speech",
    headers={"Authorization": "Bearer YOUR_DEV_KEY"},
    json={"model": "tts-1", "input": "Hello", "voice": "nova"},
)
with open("speech.mp3", "wb") as f:
    f.write(r.content)

POST https://www.myai168.com/en/api/openai/v1/audio/transcriptions

Speech-to-text (STT, multipart upload).

curl https://www.myai168.com/en/api/openai/v1/audio/transcriptions \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -F file=@audio.mp3 \
  -F model=gpt-4o-transcribe

import requests

with open("audio.mp3", "rb") as f:
    r = requests.post(
        "https://www.myai168.com/en/api/openai/v1/audio/transcriptions",
        headers={"Authorization": "Bearer YOUR_DEV_KEY"},
        files={"file": f},
        data={"model": "gpt-4o-transcribe"},
    )
print(r.json())

POST https://www.myai168.com/en/api/openai/v1/audio/translations

Translate speech to English text (STT + translate). The official OpenAI translations endpoint only supports whisper-1 (gpt-4o-transcribe is for transcriptions only); the examples pass response_format=verbose_json to get the full JSON including audio duration.

curl https://www.myai168.com/en/api/openai/v1/audio/translations \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -F file=@audio.mp3 \
  -F model=whisper-1 \
  -F response_format=verbose_json

import requests

with open("audio.mp3", "rb") as f:
    r = requests.post(
        "https://www.myai168.com/en/api/openai/v1/audio/translations",
        headers={"Authorization": "Bearer YOUR_DEV_KEY"},
        files={"file": f},
        data={"model": "whisper-1", "response_format": "verbose_json"},
    )
print(r.json())

POST https://www.myai168.com/en/api/openai/v1/images/generations

Text-to-image (gpt-image-2).

curl https://www.myai168.com/en/api/openai/v1/images/generations \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-image-2","prompt":"A cute cat","size":"1024x1024"}'

import requests

r = requests.post(
    "https://www.myai168.com/en/api/openai/v1/images/generations",
    headers={"Authorization": "Bearer YOUR_DEV_KEY"},
    json={"model": "gpt-image-2", "prompt": "A cute cat", "size": "1024x1024"},
)
print(r.json())

POST https://www.myai168.com/en/api/openai/v1/embeddings

Text embeddings (OpenAI-compatible format; accepted models voyage-4 / voyage-4-large, billed per input token).

curl https://www.myai168.com/en/api/openai/v1/embeddings \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"voyage-4-large","input":"Hello"}'

import requests

r = requests.post(
    "https://www.myai168.com/en/api/openai/v1/embeddings",
    headers={"Authorization": "Bearer YOUR_DEV_KEY"},
    json={"model": "voyage-4-large", "input": "Hello"},
)
print(r.json())

GET https://www.myai168.com/en/api/openai/v1/audio/voices

List available built-in TTS voices (query endpoint, free). Note: this GET listing is a site extension — the official OpenAI API reference only documents POST (create custom voice) and has no GET list endpoint, so do not expect it in the official SDK surface.

curl https://www.myai168.com/en/api/openai/v1/audio/voices \
  -H "Authorization: Bearer YOUR_DEV_KEY"

POST https://www.myai168.com/en/api/openai/v1/audio/voices

Custom voice creation (voice cloning, multipart upload; fields: audio_sample voice sample file required, consent consent recording ID required (e.g. cons_1234 — per the official OpenAI spec this is obtained by uploading a consent recording first, not a free-text consent statement), name voice name). OpenAI currently limits custom voices to eligible accounts.

curl https://www.myai168.com/en/api/openai/v1/audio/voices \
  -H "Authorization: Bearer YOUR_DEV_KEY" \
  -F audio_sample=@sample.mp3 \
  -F consent="cons_1234" \
  -F name="My Voice"

GET https://www.myai168.com/en/api/openai/v1/models / GET https://www.myai168.com/en/api/openai/v1/models/{model}

List supported models / retrieve one model (compatible with SDK models.list / models.retrieve; query endpoints, free).

curl https://www.myai168.com/en/api/openai/v1/models \
  -H "Authorization: Bearer YOUR_DEV_KEY"

curl https://www.myai168.com/en/api/openai/v1/models/gpt-5.6-sol \
  -H "Authorization: Bearer YOUR_DEV_KEY"

Support Matrix

Some OpenAI SDK features behave differently through this relay. Check this matrix before integrating to avoid surprises.

• GET /v1/models and GET /v1/models/{model} — supported (compatible with SDK models.list / models.retrieve). Returns exactly the models this relay accepts; models not in the list are rejected with "No pricing info! Please use other models!".
• POST /v1/responses and POST /v1/chat/completions — supported, including streaming.
• Web search — supported and billed per search: Responses API tools: [{"type": "web_search"}], Chat Completions web_search_options.
• Other hosted tools (code_interpreter, file_search, image_generation, computer use, hosted MCP) — not supported, HTTP 400. Use client-side function calling instead.
• Audio endpoints (speech / transcriptions / translations / voices — GET lists built-in voices for free, POST creates a custom voice), images/generations and embeddings (accepted models voyage-4 / voyage-4-large) — supported.
• Non-streaming responses include X-Cost-Points (points charged) and X-Credits-Remaining (balance after the call) headers. Streaming responses append the same information after the final event (data: [DONE]) as SSE comment lines : cost=N / : credits-remaining=M. You can also query your balance any time with GET https://www.myai168.com/en/api/user/get_credits.
• Failed calls, and calls where the vendor returns no usage info, are never charged.

Request parameters (POST https://www.myai168.com/en/api/openai/v1/chat/completions)

The request body is JSON, identical to the official OpenAI spec. Authenticate with an Authorization: Bearer YOUR_API_KEY header (never place keys in URL query strings).

The Responses API (POST /v1/responses) uses input (string or array, required) and model (required) instead, and accepts tools: [{"type": "web_search"}]; other official parameters are forwarded as-is.

Response format

Non-streaming: official JSON — Chat Completions text is in choices[0].message.content with usage in usage; Responses text must be parsed from the output[] array (in items whose type is message, take the text field of content[] blocks whose type is output_text) — output_text is a convenience property of the official Python / JavaScript SDKs, not a field of the raw REST JSON. Response headers include X-Cost-Points (points charged) and X-Credits-Remaining (balance after the call).

Streaming: SSE chunks with Chat Completions deltas in choices[0].delta.content, ending with data: [DONE]; after the final event the stream appends SSE comment lines : cost=N / : credits-remaining=M with the points charged and remaining balance (spec-compliant SSE parsers and official SDKs ignore comment lines). You can also query GET https://www.myai168.com/en/api/user/get_credits any time.

Streaming example

curl -N https://www.myai168.com/en/api/openai/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-sol","stream":true,"messages":[{"role":"user","content":"Hello"}]}'

data: {"id":"...","object":"chat.completion.chunk","choices":[{"delta":{"content":"Hi"},"index":0}]}

data: {"id":"...","object":"chat.completion.chunk","choices":[{"delta":{},"finish_reason":"stop","index":0}]}

data: [DONE]

Error codes

No points are ever charged on errors (including when the vendor returns no usage info).