mudler/LocalAI
LocalAI is an open-source AI engine that enables running various AI models (LLMs, vision, voice, image, video) on any hardware, including CPU-only systems, with a composable and extensible architec...
Awesome AI Agents › LLM Framework Tools
MoonshotAI/kosong
Kosong is a Python library published by Moonshot AI that acts as an abstraction layer between agent application code and the underlying large language model. It is the foundation used by Moonshot's own Kimi CLI agent, and it exposes a provider-neutral surface for chat completion, streaming, tool or function calling and message construction, so that an agent loop can be written once and pointed at Kimi models or other compatible backends without rewriting the transport and schema handling. The package targets the layer that most agent projects end up reimplementing: normalising request and response shapes across providers, describing tools to the model, parsing tool calls back out of a completion, and streaming partial output while the call is still in flight. The repository at MoonshotAI/kosong is now a pointer rather than the working tree. Its README states that development has moved to the kimi-cli monorepo, where the code is kept under packages/kosong, so anyone adopting the library should install and track it from that location and treat this repository as historical context and as the address most references still point at. That relocation matters when judging the project: issues, releases and commits happen in the monorepo, and the standalone repository will not show current activity even though the library itself is maintained. Kosong is aimed at developers building agents on Moonshot's Kimi family of models, and more generally at anyone who wants a small, single-purpose LLM client layer underneath their own agent runtime instead of a large opinionated framework that also dictates memory, planning and orchestration. Read it as an SDK component, not as a complete agent framework.
https://github.com/MoonshotAI/kosong
Also in LLM Framework Tools
langfuse/langfuse
Langfuse is an open source LLM engineering platform that enables collaborative development, monitoring, evaluation, and debugging of AI applications with comprehensive observability, prompt management, and integration features.
567-labs/instructor
Instructor is a popular Python library that simplifies working with structured outputs from large language models by providing validation, retries, and streaming support across multiple LLM providers.
TEN-framework/ten-framework
TEN Framework is a real-time, distributed, cloud-edge collaborative multimodal AI agent framework supporting multiple programming languages for building high-performance AI applications.
google/adk-samples
A collection of sample agents built with the Agent Development Kit (ADK) to accelerate development of conversational and multi-agent workflows, designed for customization and integration with Google Cloud services.
truefoundry/cognita
Cognita is an open-source, modular, and production-ready RAG framework by TrueFoundry that simplifies building, customizing, and deploying scalable Retrieval Augmented Generation applications with a no-code UI and extensive integration features.
lemony-ai/cascadeflow
In-process Python and TypeScript library that cascades model calls inside the agent loop, routing each query or tool call between small and flagship models by quality, cost, latency and budget.
shroominic/codeinterpreter-api
An open-source API that implements the ChatGPT Code Interpreter for sandboxed Python code execution, enabling data analysis, visualization, and interactive AI-driven computational tasks with local execution and scalable production deployment.