Release v0.1.9
FlexLLama v0.1.9 Release Notes
Overview
This release adds support for the OpenAI Responses API, enabling compatibility with the latest OpenAI client libraries and agent frameworks that use the /v1/responses endpoint.
馃殌 New Features
OpenAI Responses API Support
Added: New /v1/responses endpoint that proxies requests to the underlying llama.cpp server. Supports both streaming and non-streaming modes, system-level instructions, and model routing with automatic default model fallback. Requires a llama.cpp build with Responses API support (PR #18227+).
Impact: Users can now interact with FlexLLama using the OpenAI Responses API format, unlocking compatibility with agent frameworks and newer OpenAI SDKs that rely on this endpoint.
Bring your agents to local inference with FlexLLama v0.1.9!