OpenAICompatibleHTTPLLMClient

OpenAICompatibleHTTPLLMClient targets any OpenAI-compatible HTTP endpoint.

Default behavior

  • Default base URL: http://127.0.0.1:8001/v1

  • Default model: qwen2-1.5b-q4

  • Execution locality depends on where your compatible endpoint is hosted

Constructor-first usage

from design_research_agents import OpenAICompatibleHTTPLLMClient
from design_research_agents.llm import LLMMessage, LLMRequest

client = OpenAICompatibleHTTPLLMClient(base_url="http://127.0.0.1:8001/v1")
response = client.generate(
    LLMRequest(
        messages=(LLMMessage(role="user", content="Draft a one-line abstract."),),
        model=client.default_model(),
    )
)

Dependencies and environment

  • Reachable OpenAI-compatible HTTP endpoint

  • Optional API key env variable depending on upstream server behavior

Examples

  • examples/clients/openai_compatible_http_client.py

Attribution