Skip to main content
Everything you need to hit the core endpoints in one place: text generation, image generation, and audio transcription.

​
Text generation

import openai

client = openai.OpenAI(
    base_url="https://api.llm7.io/v1",
    api_key="none",  # or your token
)

resp = client.chat.completions.create(
    model="default",  # or "fast" / "pro"
    messages=[
        {"role": "system", "content": "Answer concisely."},
        {"role": "user", "content": "Give me three onboarding tips for new engineers."},
    ],
    temperature=0.4,
)

print(resp.choices[0].message.content)
Use model="fast" for lowest latency and model="pro" (paid) for the highest quality reasoning.

​
Image generation

from openai import OpenAI

client = OpenAI(base_url="https://api.llm7.io/v1", api_key="none")

prompt = (
    "A futuristic cityscape at sunset, flying cars, neon reflections, "
    "cinematic cyberpunk, highly detailed"
)

res = client.images.generate(
    model="flux",          # or "turbo" (aliases: 1, 2, image-model-1, image-model-2)
    prompt=prompt,
    size="1024x1024",      # or set w / h (100–1500)
    extra_body={"seed": 42},
)

print(res.data[0].url)

​
Audio transcription

import openai
import pathlib

client = openai.OpenAI(
    base_url="https://api.llm7.io/v1",
    api_key="none",  # or your token
)

def transcribe(audio_path: str):
    with pathlib.Path(audio_path).open("rb") as fp:
        return client.audio.transcriptions.create(
            model="gpt-4o-mini-audio-preview",
            file=fp,
            language="en",
            response_format="json",
            temperature=0,
        )

print(transcribe("1.mp3"))
Expected response structure: 
Transcription(text="Transcription...", logprobs=None)
Audio supports .mp3 and .wav. Set language to improve accuracy for non-English audio.