SuperAPI AI Directory
Partner with us中文AI newsFavorites

All AI tools

35,738 results · Page 797 of 993

OpenHands

AI agents

OpenHands is a self-hosted developer control center for managing and running coding agents and automation tasks. It supports running OpenHands, Claude Code, Codex, Gemini, Pi, OpenCode, or any ACP-compatible agent on local, remote, and cloud backends. Through OpenHands Agent Canvas, users can organize coding agents into an always-on engineering team to initiate conversations and automate routine tasks, such as generating reports and posting them to Slack, or automatically breaking down GitHub issues into tasks. Agent Canvas runs on the local machine by default, and can also connect to multiple agent backends, such as Docker containers, virtual machines, or company infrastructure, with optional use of OpenHands Cloud or OpenHands Enterprise.

Read more

OpenLane

Other AI tools

Lane dataset. OpenLane is the first real-world and largest 3D lane dataset to date. Our dataset collects valuable content from the public perception dataset Waymo Open Dataset and provides lane and closest in-path object (CIPO) annotations for 1000 road segments. In short, OpenLane has 200K frames and over 880K carefully annotated lanes. We publicly released the OpenLane dataset to help the research community make progress in 3D perception and autonomous driving technology.

Read more

OpenManus-RL

AI developer platforms

OpenManus-RL is an open-source project jointly led by Ulab-UIUC and MetaGPT. It is an extended version of OpenManus, focused on reinforcement learning (RL) tuning for large language model (LLM) agents. The project explores a new paradigm for RL-based LLM agent training, drawing on the successful experience of reasoning models such as Deepseek-R1 and QwQ-32B in RL tuning, and regularly updates exploration directions and results through livestream-style development. The project conducts rigorous testing on agent benchmarks such as GAIA, AgentBench, WebShop, and OSWorld, and publicly shares the tuned models. The code and datasets have been released, and the verl submodule has been integrated to enhance RL training capabilities. The project supports agent environment setup, trajectory data collection, RL fine-tuning paradigms, and benchmark evaluation, and welcomes community contributions of code, datasets, environments, or compute resources.

Read more

OpenObserve

Other AI tools

OpenObserve is an open-source, unified observability platform for the AI era, designed for telemetry data such as logs, metrics, and traces. It is written in Rust, based on the DataFusion engine and columnar Parquet storage, and the official description states that no indexes need to be built, which can significantly reduce storage and compute costs (claimed storage costs are about 1/140 of Elasticsearch and compute about 1/30), supporting PB-scale log retrieval and Prometheus-compatible metric queries. The platform provides dashboards, alerts, SLOs, RUM, synthetic monitoring, AI SRE agents, AI assistant (natural language queries), LLM observability, and other capabilities, and can automatically correlate logs, metrics, and traces. It supports native OpenTelemetry ingestion, one-click collector deployment, and has Kubernetes container log, cluster event, and metric collection as well as zero-code tracing capabilities. The official description states that it already has more than 10,000 enterprises and 22.3k stars, emphasizing no vendor lock-in, with data stored in the user's own bucket.

Read more

openpaw

AI agents

OpenPaw is a personal assistant configuration wizard for Claude Code. With a single npx pawmode command, it extends Claude Code, which can only write code, into a personal assistant capable of handling tasks such as email, calendar, music, smart home, Slack, GitHub, Obsidian, and more. It installs 38 skills and their corresponding CLI tools, guides you through authentication, configures an identity and persistent memory for Claude, sets up permissions and security hooks, connects a Telegram bridge, provides a local kanban task panel, supports scheduled tasks with cost caps, and optionally enables self-learning, token compression, and theme beautification. Everything runs locally, with no daemon, cloud service, or subscription required, reusing your existing Claude Code subscription.

Read more

OpenPilot

Other AI tools

OpenPilot is an open-source AI agent runtime platform (agent harness) that runs as a local Electron desktop application on laptops. Users can bring their own LLM API key for any OpenAI-compatible interface, and can optionally integrate TinyFish to enable real-time web search, allowing the agent to read, edit, and invoke tools on real project folders, with no vendor lock-in, and keys kept on the local machine. The project uses the MIT license, provides prebuilt installers on GitHub Releases, and supports automatic updates.

Read more

Opensource Boilerplates

Other AI tools

OpenSourceBoilerplates is a hand-picked collection site of open-source boilerplates and SaaS starter templates, designed to help developers quickly start new projects. Its coverage spans multiple tech stacks, including React, Next.js, Node.js, Python, Django, Flutter, React Native, and native platforms, and includes both web application templates and mobile app templates. The site lists specific repository names, such as android/architecture-samples, fastapi/full-stack-fastapi-template, sahat/hackathon-starter, infinitered/ignite, nextjs/saas-starter, wasp-lang/open-saas, and others. In addition, the page states that a sponsored display spot is available, contactable via X. It should be noted that, apart from collecting and displaying listings, the site itself currently does not provide other明确 features, fee information, or pricing details.

Read more

OpenVids

AI video

OpenVids is an open-source, Agent-centric AI video editor that supports macOS and Windows. Users only need to describe their film requirements in natural language, and the Agent will automatically build chapters, edit the timeline, add motion effects and sound, and review the rendered result. It is aimed at professional editors, emphasizes keyboard operation, non-destructive editing, and media files stored on the local disk, with the final decision-making power still belonging to the user.

Read more