Search

bb helper inference (thread titles, commit messages) over OpenAI Chat Completions, for a LiteLLM gateway or a local server.

Restricted Telegram personal-account history, operator replies and reaction copying

A framework-agnostic Python library for classifying LLM prompts and routing requests to the best-fit model and provider.

Benchmark Apple Silicon memory bandwidth, disk speed, and MLX model inference performance to find the optimal LLM size

TypeScript client for the AnswerLine answer and search monitoring API.

Python SDK for LuxurAI — 322 Neural Voice Models, LLM Streaming, Unified History, Chat & Code

国内大模型选择器/路由器：自然语言任务分类 × 评分矩阵推荐 × OpenAI 兼容网关路由

n8n node for Statable, privacy-first cookieless web analytics. Query visitors, sources, pages, goals and funnels in your workflows.

Built-in OpenAI-compatible model provider for Alisio.

The LLM gateway that learns to skip the LLM: trains its own tiny decision model on your agent's traffic and answers repeat decisions in 0.2 ms with no LLM call. OpenAI-compatible, one Rust binary. `npx pankhllm serve`.

Fast, dependency-light Telegram Bot API framework built on urllib3

Apple's on-device (Foundation Models) and Private Cloud Compute LLMs from Python. No API key, no account, macOS 26+ on Apple Silicon.

Python SDK for Kinesis Video Webrtc Storage.

Vercel AI SDK provider for NuRoute — route generateText/streamText through your NuRoute gateway

n8n node for DaMeng (达梦) Database - 支持 Execute SQL、Select、Insert、Update、Delete 操作

Test a RAG fix against your real retriever before you ship it.

Interchange ProviderAdapter for the OpenAI Responses API wire protocol with quirks-based vendor configuration, SSE and non-streaming parsing, and reasoning signature replay.

Thin LangChain integration for Komplex AI — a hallucination detector for LLM output (adapts the halu SDK to LCEL).

Generate and edit video, image and audio with Magic Hour

Use the Whatsplaid API in n8n workflows.