yetter 0.0.5

Yetter Python Client

Yetter Python Client

A lightweight async Python client for the Yetter image generation API. It supports simple one-shot runs, subscription-style polling, and server-sent events (SSE) streaming.

Installation

pip install yetter

If the package is not yet published to PyPI, install from the repository:

pip install "yetter @ git+https://github.com/SqueezeBits/yetter-client"

Authentication & Configuration

Provide your API key in one of the following ways:

• Environment variable (preferred):

export YTR_API_KEY="YOUR_API_KEY"

• Programmatically via configure:

import asyncio
import yetter

async def main():
    yetter.configure(api_key="YOUR_API_KEY")
    # Optionally override endpoint:
    # yetter.configure(api_key="YOUR_API_KEY", api_endpoint="https://api.yetter.ai")

asyncio.run(main())

Quick start: one-shot run

Use yetter.run() for a convenient "submit + wait for completion" flow powered by streaming under the hood.

import asyncio
import yetter

async def main():
    yetter.configure(api_key="YOUR_API_KEY")
    result = await yetter.run(
        "ytr-ai/qwen/image/t2i",
        args={"prompt": "A beautiful landscape with a river and mountains"},
    )
    # result is dict('images':..., 'prompt':...)
    print(result)

asyncio.run(main())

Subscribe: polling with optional queue updates

yetter.subscribe() submits a request and polls status until completion. You can pass a callback to receive queue/status updates.

import asyncio
import yetter
from yetter.types import GetStatusResponse

async def on_update(status: GetStatusResponse):
    print("status:", status.status)

async def main():
    yetter.configure(api_key="YOUR_API_KEY")
    result = await yetter.subscribe(
        "ytr-ai/qwen/image/t2i",
        args={
            "prompt": "A beautiful landscape with a river and mountains",
            "num_images": 2,
        },
        on_queue_update=on_update,
    )
    print("final images:", result.images)

asyncio.run(main())

File Upload

Upload files to Yetter using presigned URLs. Supports both single-part upload for small files and automatic multipart upload for large files.

import asyncio
import yetter

async def main():
    yetter.configure(api_key="YOUR_API_KEY")

    # Simple upload
    result = await yetter.upload_file("./my_image.jpg")
    print(f"Uploaded: {result.url}")

    # Upload with progress tracking
    def on_progress(percent: int):
        print(f"Upload progress: {percent}%")

    result = await yetter.upload_file(
        "./large_file.png",
        on_progress=on_progress,
    )
    print(f"Uploaded: {result.url}")

asyncio.run(main())

The upload_file function returns an UploadCompleteResponse containing:

• url: Public URL to access the uploaded file
• key: S3 object key for reference
• metadata: Optional metadata (size, content_type, uploaded_at)

Examples

See the examples/ directory:

• examples/run.py: minimal one-shot run
• examples/subscribe.py: subscription with updates

Development Notes

When contributing to this repository, use TDD by default:

• add or update a failing test first
• implement the smallest change that makes it pass
• refactor only after the test suite is green again

Recommended local test command:

uv run --extra dev pytest -q