Lower-visibility AI repositories with strong quality and maintenance signals.
Self-hosted AI agent OS. Your memory, chat, agents, and files stay on hardware you own, offline by default, cloud by choice. Offline AI memory (taOSmd), self-hosted multi-framework group chat, a full
Why: +194 stars in 30 days; 100 commits in 30 days
agent-framework ai-agents ai-platform apple-silicon data-sovereignty distributed-computing
Unified LangChain documentation.
Why: +5 stars in 7 days; 100 commits in 30 days
documentation langchain langgraph langsmith
RDNA-native LLM inference engine in Rust.
Why: +8 stars in 7 days; 100 commits in 30 days
amd-gpu gpu-computing hip llm-inference machine-learning quantization
Engine-agnostic LLM gateway in Rust. Full OpenAI & Anthropic API compatibility across vLLM, TRT-LLM, TokenSpeed, SGLang, OpenAI, Gemini & more. Industry-first gRPC pipeline, KV cache-aware routing, ch
Why: +10 stars in 7 days; 100 commits in 30 days
anthropic anthropic-api chat claude gemini inference-gateway
Auditable workspaces for AI coding agents: sandboxed worktrees, conflict-free multi-agent orchestra, 95% lower token waste, and persistent memory.
Why: +10 stars in 7 days; 100 commits in 30 days
agent agentic-ai agentic-workflow ai-coding ai-safety ai-security
Aurora — Open source AI-powered agentic incident management & root cause analysis for SREs. LangGraph agents investigate across AWS, Azure, GCP, Kubernetes. Integrates with PagerDuty, Datadog, Grafana
Why: +7 stars in 7 days; 56 commits in 30 days
ai-agents aiops automation aws azure datadog
LLM-supervised persistent memory for AI agents — graph-based recall, cross-session knowledge, single binary. Works with Claude Code, OpenClaw, and any CLI agent.
Why: +5 stars in 7 days; 100 commits in 30 days
agent-framework agent-memory ai-agent ai-tools claude claude-code
Design, conduct and analyze results of AI-powered surveys and experiments. Simulate social science and market research with large numbers of AI agents and LLMs.
Why: +3 stars in 7 days; 95 commits in 30 days
anthropic data-labeling deepinfra domain-specific-language experiments llama2
LvLLM is a special NUMA extension of vllm that makes full use of CPU and memory resources, reduces GPU memory requirements, and features an efficient GPU parallel and NUMA parallel architecture, suppo
Why: +4 stars in 7 days; 100 commits in 30 days
cpu decode gpu hybrid inference model
Hyperparameter optimization and feature selection for scikit-learn using evolutionary algorithms. A modern alternative to GridSearchCV and RandomizedSearchCV.
Why: +15 stars in 7 days; 100 commits in 30 days
artificial-intelligence automl cross-validation evolutionary-algorithms feature-engineering feature-selection
The memory your AI should have had from the start. Automatic capture, automatic recall, 100% local. One SQLite file, zero cloud. Works with Claude Code, Claude CLI, Cursor, Codex CLI, Gemini CLI.
Why: +3 stars in 7 days; 100 commits in 30 days
agent-memory ai ai-agent ai-agents ai-memory anthropic
The open creative AI workspace
Why: +8 stars in 7 days; 100 commits in 30 days
anthropic cerebras cohere deepseek fal gemini
Runtime policy enforcement for AI agents. Cryptographic audit trail, human-in-the-loop approvals, kill switch. Zero code changes.
Why: +3 stars in 7 days; 85 commits in 30 days
ai-agents ai-safety anthropic audit-trail langchain llm-observability
Local AI Assistant on your phone
Why: +11 stars in 7 days; 89 commits in 30 days
ai gemma3 gemma3n gemma4 gemma4-agent-skills gptoss
Symbol Delta Ledger (SDL-MCP) is a policy-centered context budget layer for coding agents: Symbol-graph intelligence combined with precision tools. It turns sprawling codebases into compact, high-sig
Why: +13 stars in 7 days; 100 commits in 30 days
agent-context agent-tools agentic-coding agentic-engineering agentic-workflow code-analysis
Build production-ready backends in minutes
Why: +1 stars in 7 days; 34 commits in 30 days
access-control api authorization awslambda cloudflare-workers deno
Open Model Engine (OME) — Kubernetes operator for LLM serving, GPU scheduling, and model lifecycle management. Works with SGLang, vLLM, TensorRT-LLM, and Triton
Why: +5 stars in 7 days; 23 commits in 30 days
deepseek k8s kimi-k2 llama llm llm-inference
KVarN is a native vLLM KV-cache quantization backend for your agents: 3-5x more context, throughput above FP16, and FP16-level accuracy. Calibration-free, one flag.
Why: +10 stars in 7 days; 100 commits in 30 days
agentic-ai kv-cache llm llm-inference long-context quantization
Explore the unknown, build the future, own your data.
Why: +3 stars in 7 days; 100 commits in 30 days
agents ai ai-agents artificial-intelligence ide llm-inference
Open-source runtime AI agent security tool - monitors and controls AI agents, catching malicious tool use, prompt injection, and policy drift in real time, before the agent acts.
Why: +21 stars in 7 days; 19 commits in 30 days
aarm agent-security agentic-ai agentic-security agents ai
Production-grade Go SDK for building AI agents with long-term memory, knowledge retrieval, and voice — runnable as a library, a daemon, or a real-time pipeline.
Why: +1 stars in 7 days; 29 commits in 30 days
agent-framework agent-runtime agents ai ai-agents anthropic
Build computer vision models in a fraction of the time and with less data.
Why: 10 commits in 30 days; 35 active contributors
computer-vision deep-learning fine-tuning geti inference openvino
Krasis is a Hybrid LLM runtime which focuses on efficient running of larger models on consumer grade VRAM limited hardware
Why: +3 stars in 7 days; 40 commits in 30 days
cpu-inference gguf-model-support gpu-inference high-performance-inference hybrid-inference inference-engine
sparkrun - launch, manage, and stop LLM inference workloads on NVIDIA DGX Spark systems
Why: +15 stars in 7 days; 13 commits in 30 days
dgx-spark inference llama-cpp sglang vllm
Redis Vector Library (RedisVL) -- the AI-native Python client for Redis.
Why: +2 stars in 7 days; 19 commits in 30 days
anthropic embedding huggingface large-language-models llm llmcache
Program that lets you ask questions about your documents, audio, and video files.
Why: +1 stars in 7 days; 49 commits in 30 days
bark database-management embedding-models embedding-vectors embeddings gtts
The Runpod worker template for serving our large language model endpoints. Powered by vLLM.
Why: +3 stars in 30 days; 27 commits in 30 days
language-model llm runpod vllm
An Open-source, self-hosted AI model hub with Hugging Face compatibility, accelerating vLLM/SGLang performance.
Why: +3 stars in 7 days
artificial-intelligence dynamo huggingface kubernetes llm llm-inference
Vector search engine inside Milvus, integrating FAISS, HNSW, DiskANN.
Why: +3 stars in 7 days; 14 commits in 30 days
faiss gpu hnsw nearest-neighbor-search search vector
Python package for LLM compression
Why: +3 stars in 7 days; 10 active contributors
llm qep quantization vllm