promtexpress

Official Python client for the PromtExpress API. Standard library only, Python 3.9+.

pip install promtexpress

Usage

from promtexpress import PromtExpress

client = PromtExpress()  # reads PROMTEXPRESS_API_KEY

result = client.generate(
    "product launch email for a note-taking app, friendly tone",
    "text",  # text | code | image | video | audio | music
)
print(result["output"])
print(result["creditsRemaining"], "credits left")

Clarifying questions and iteration

if result.get("chipQuestions"):
    better = client.generate(
        "product launch email for a note-taking app",
        "text",
        answers=[{"question": result["chipQuestions"][0]["label"], "answer": "developers"}],
    )

refined = client.generate(
    "product launch email for a note-taking app",
    "text",
    iteration={"ofPromptId": result["promptId"], "feedback": "shorter, mention the free tier"},
)

Templates and history

templates = client.list_templates(modality="image")

page = client.list_history(page=0, limit=50)

for entry in client.iter_history(modality="video"):
    print(entry["date"], entry["title"])

API key scopes

Create keys in the PromtExpress dashboard under API Keys. The read scope covers list_templates, list_history and iter_history; generate covers generate (consumes credits); admin allows everything. Each key has its own rate limit.

Errors

All errors inherit from PromtExpressError: InvalidRequestError (400), AuthenticationError (401), PermissionDeniedError (403, with missing_scope), InsufficientCreditsError (402, with remaining and required), RateLimitError (429, with retry_after), ServerError (5xx) and APIConnectionError (network or timeout).

Time-based rate limits are retried automatically (max_retries, default 2).

License

MIT