SuperAPI AI Directory
Partner with us中文AI newsFavorites

All AI tools

35,738 results · Page 325 of 993

openhanako

AI design

HanaAgent (repository name openhanako) is a personal AI assistant designed for ordinary computer users, with memory, personality, and autonomous action capabilities. It does not require users to write complex configurations or understand obscure terminology, and provides a relatively complete graphical interface rather than being limited to the command line. It can remember things the user has said, operate the user's computer, browse web pages, search for information, read and write files, execute code, manage schedules, and autonomously learn new skills; it supports multiple Agents collaborating on the same computer, with each Agent having independent memory, personality, and scheduled tasks. Agents can collaborate through channel group chats and delegate tasks to each other. It also provides a desk, full-screen media viewer, session management, scheduled tasks and heartbeat, secure sandbox, plugin system, and other features, and supports integration with Telegram, Feishu, QQ, WeChat bots, as well as mobile PWA and LAN frontends. The interface supports 5 languages: Chinese, English, Japanese, Korean, and Traditional Chinese. It provides installation packages for macOS, Windows, and Linux, and is licensed under the Apache License 2.0.

Read more

OpenHands

AI coding

OpenHands is a self-hosted developer control center for running coding agents and automating tasks. It supports running OpenHands, Claude Code, Codex, Gemini, Pi, OpenCode, or any ACP-compatible agent, and can connect to local, remote, and cloud backends. Its Agent Canvas interface turns coding agents into a persistent engineering team, used to start conversations and automate everyday tasks, such as generating reports and posting them to Slack, or automatically breaking down GitHub issues into tasks. It runs locally by default, and can also connect to agent backends in Docker containers, virtual machines, or company infrastructure, and can optionally use OpenHands Cloud or OpenHands Enterprise infrastructure.

Read more

openhermit

AI developer platforms

OpenHermit is an open-source AI agent deployment platform built by Amiko, used to run large numbers of AI agents as production-grade services. Its core design separates internal state from external state: internal state such as sessions, memory, instructions, skills, MCP servers, scheduled tasks, secrets, and users is stored centrally in a shared PostgreSQL and isolated by agent_id scope; external state such as the workspace files an agent is currently working on is placed in each agent's independent sandbox, with optional backends such as self-hosted Docker, E2B, Daytona, or Tenki. The platform provides a control plane through a single Hono gateway server; agents can be started, attached, and detached at any time, and it includes a built-in /admin/ management interface. It has built-in Telegram, Discord, and Slack channel adapters, as well as Signal, WeChat, and WhatsApp adapters installable via packages, plus a CLI and Web UI, supporting runtime enablement, disabling, and reconfiguration. Skills and MCP servers can be centrally installed and enabled per agent or across the entire fleet. It also provides cron and one-off scheduled tasks with timeouts, concurrency policies, and error backoff, multi-user roles (owner / user / guest), multi-protocol transport (HTTP synchronous, inline SSE, persistent SSE, WebSocket RPC), and a deep research feature that cites server-verified evidence. The project uses the MIT license and is still under continuous development.

Read more

OpenJudge Review

AI search

OpenJudge Review is a paper review-related service provided by OpenJudge. According to the official statement, its web-based literature review service has been discontinued, and users are temporarily unable to upload PDFs or retrieve review reports on the web. The web-based paper review, Rebuttal generation, and evaluation functions have all ceased to be offered. The official recommendation is to switch to Codex and use OpenJudge's paper review skills to complete review work locally. These skills cover safety checks, correctness evaluation, quality scoring, critical analysis, and reference verification. They can process PDF or TeX files locally and output review reports in both Chinese and English. The specific web-based pricing, usage conditions, and restoration time have not yet been provided.

Read more

openlens-ai

AI agents

OpenLens AI is a fully autonomous multimodal scientific research agent based on the LangGraph multi-module architecture, capable of automatically completing scientific research workflows such as literature review, data analysis, experimental design, code generation and execution, and LaTeX paper generation. It provides a Streamlit interactive interface, supports Chinese paper writing, and can execute code through the OpenHands sandbox. The project is open source, uses the MIT License, requires self-configuration of LLM and Tavily search API keys, and depends on a Docker runtime environment.

Read more

Openmart

AI developer platforms

Openmart is a business data platform for local businesses and small and medium-sized enterprises (SMBs), offering a local merchant database, API, and data enrichment capabilities. Users can search for local businesses by industry and region, view merchant profiles, websites, public contact information, ratings, and market insights for market research, lead generation, and sales outreach. The platform claims to cover more than 500 local business categories and 16 markets, with a database containing over 50 million merchants. Its AI sales agent, Openclaw, can automatically generate an execution plan based on the user's sales goals, find and score leads, and draft personalized outreach emails, outputting structured lead lists and exportable files. Openmart also supports integration with Clay and states that it complies with data protection regulations such as GDPR and CCPA. Specific pricing and actual results are not yet clearly provided and should be subject to official channels.

Read more

OpenOmni

AI developer platforms

OpenOmni is an open-source omnimodal large language model project, and the related paper has been accepted by NIPS 2025. Through progressive multimodal alignment and real-time self-aware emotional speech synthesis technology, the project enables input and output across multiple modalities such as speech, text, images, and video. The repository provides model code, pretrained weights, data, and training and evaluation scripts, supports fast inference and interaction with omnimodal input, and includes complete pipelines such as speech-to-text pretraining, image-to-text instruction fine-tuning, text-to-speech pretraining, and emotional DPO tuning. The project also provides multiple benchmark evaluations, such as Speech-Text, Image-Text, Speech-Text-Image, and Speech-Text-Image-Video, along with examples of Chinese tongue twisters and emotional speech synthesis. It is suitable for developers and researchers studying multimodal large models, speech interaction, and emotional speech synthesis. Some specific features and pricing information are not yet available.

Read more

openr

AI developer platforms

OpenR is an open-source framework for advanced reasoning with large language models, maintained by the Openreasoner team and released under the MIT License. The project provides resources such as papers, tutorials, code, documentation, data, models, issue tracking, and demos, and has released the MATH-APS dataset and the MATH-psa process reward model. OpenR supports MCTS reasoning and provides benchmarks. You first need to download the specified base models (such as Qwen2.5-Math-1.5B-Instruct, Qwen2.5-Math-7B-Instruct, peiyi9979/mistral-7b-sft, peiyi9979/math-shepherd-mistral-7b-prm), and configure the model path, policy model name, value model name, and the number of LM/RM workers by modifying the scripts under reason/llm_service/. After that, you can prepare and run reasoning with different techniques. Before training, you need to modify the dataset path, model path, and PRM path in train/mat/scripts/train_llm.sh. The project welcomes community contributions, not limited to code, but also including answering questions, improving documentation, and more.

Read more

OpenRFT

AI developer platforms

OpenRFT is an open-source project that aims to adapt general reasoning foundation models to domain-specific tasks through Reinforcement Fine-Tuning (RFT). The project leverages domain-specific samples to achieve efficient fine-tuning for domain-specific tasks, addressing challenges such as the lack of reasoning step data and the limited number of training samples. The project's training code relies on the OpenRLHF and trl frameworks and is released under the MIT License.

Read more