amyrmahdy-llmwrap 0.1.0

Lightweight, production-ready adapter for LLM chat completions (text + multimodal)

llmwrap

Lightweight adapter for LLM chat completions — text, images, PDFs/files.
OpenAI-compatible (works great with OpenRouter, Anthropic via compat layer, etc.).

Installation

pip install llmwrap
# or
uv add llmwrap

Quick start

from llmwrap import LLM

llm = LLM(
    api_key="sk-...",
    base_url="https://openrouter.ai/api/v1",
    model="anthropic/claude-3.5-sonnet",
)

# Text
print(llm.complete([{"role": "user", "content": "Hi!"}]))

# Image + text
messages = [
    {"role": "user", "content": [
        LLM.text_content("Describe this."),
        LLM.image_content("photo.jpg"),
    ]}
]
print(llm.complete(messages))