Business Software for LM Studio

OpenAI’s mission is to ensure that artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity. We will attempt to directly build safe and beneficial AGI, but will also consider our mission fulfilled if our work aids others to achieve this outcome. Apply our API to any language task — semantic search, summarization, sentiment analysis, content generation, translation, and more — with only a few examples or by specifying your task in English. One simple integration gives you access to our constantly-improving AI technology. Explore how you integrate with the API with these sample completions.

bolt.diy is an open-source platform that enables developers to easily create, run, edit, and deploy full-stack web applications with a variety of large language models (LLMs). It supports a wide range of models, including OpenAI, Anthropic, Ollama, OpenRouter, Gemini, LMStudio, Mistral, xAI, HuggingFace, DeepSeek, and Groq. The platform offers seamless integration through the Vercel AI SDK, allowing users to customize and extend their applications with the LLMs of their choice. With its intuitive interface, bolt.diy is designed to simplify AI development workflows, making it a great tool for both experimentation and production-ready applications.

Hugging Face is a leading platform for AI and machine learning, offering a vast hub for models, datasets, and tools for natural language processing (NLP) and beyond. The platform supports a wide range of applications, from text, image, and audio to 3D data analysis. Hugging Face fosters collaboration among researchers, developers, and companies by providing open-source tools like Transformers, Diffusers, and Tokenizers. It enables users to build, share, and access pre-trained models, accelerating AI development for a variety of industries.

Vicuna-13B is an open-source chatbot trained by fine-tuning LLaMA on user-shared conversations collected from ShareGPT. Preliminary evaluation using GPT-4 as a judge shows Vicuna-13B achieves more than 90%* quality of OpenAI ChatGPT and Google Bard while outperforming other models like LLaMA and Stanford Alpaca in more than 90%* of cases. The cost of training Vicuna-13B is around $300. The code and weights, along with an online demo, are publicly available for non-commercial use.

Nelly is a comprehensive AI agent platform that empowers users to build, test, distribute, and utilize AI agents without any coding required. Through Nelly Studio, users can create custom AI agents using natural language instructions, formatting them with headings, lists, and other content types. These agents can be equipped with various tools, such as a browser and a database, to accomplish their tasks. Complex tasks can be broken down into smaller problems and delegated to specialized sub-agents, allowing users to build a team of agents to handle intricate workflows. With Nelly, users can have natural, flowing conversations with their AI agents, which understand context and maintain coherent dialogue, eliminating the need for special commands or syntax. Conversations are organized into threads for better performance and organization. Users can also create departments and organize their agents using drag and drop, building their ideal AI team.

Devstral is an open source, agentic large language model (LLM) developed by Mistral AI in collaboration with All Hands AI, specifically designed for software engineering tasks. It excels at navigating complex codebases, editing multiple files, and resolving real-world issues, outperforming all open source models on the SWE-Bench Verified benchmark with a score of 46.8%. Devstral is fine-tuned from Mistral-Small-3.1 and features a long context window of up to 128,000 tokens. It is optimized for local deployment on high-end hardware, such as a Mac with 32GB RAM or an Nvidia RTX 4090 GPU, and is compatible with inference frameworks like vLLM, Transformers, and Ollama. Released under the Apache 2.0 license, Devstral is available for free and can be accessed via Hugging Face, Ollama, Kaggle, Unsloth, and LM Studio.

Broxi AI is a no-code AI agent builder that lets you go from a simple text description to a fully functional AI agent in minutes using visual, drag-and-drop tools, requiring zero technical setup. Its Broxi Autopilot takes your natural‑language briefing, such as “an agent that answers FAQs from our PDF handbook”, then lets you define input types (PDFs, chats, websites, etc.) and outputs (emails, messages, API actions). With a single click, Broxi automatically constructs, tests in an interactive playground, and enables instant deployment of your agent via API, web widget, Slack, or embedded app. It supports integrations with dozens of tools and systems, offers real-time monitoring and centralized management, and maintains enterprise-grade security throughout, making it easy for non-technical teams to automate customer support, internal workflows, sales engagement, content generation, and data retrieval, without writing a line of code.

Crush is a glamorous AI coding agent that lives right in your terminal, seamlessly connecting your tools, code, and workflows with any Large Language Model (LLM) of your choice. It offers multi-model flexibility, letting you choose from a variety of LLMs or add your own using OpenAI or Anthropic-compatible APIs, and supports mid-session switching between them while preserving context. Crush is session-based, enabling multiple project-specific contexts to coexist. Powered by Language Server Protocol (LSP) enhancements, it incorporates coding-aware context just like a developer’s editor. It's highly extensible via Model Context Protocol (MCP) plugins using HTTP, stdio, or SSE for added capabilities. Crush runs anywhere, leveraging Charm’s sleek Bubble Tea-based TUI for a polished terminal user experience. Written in Go and MIT-licensed (with FSL-1.1 for trademarks), enabling developers to stay in their terminal while taking advantage of expressive AI coding assistance.

Novelcrafter is an AI-powered writing platform designed to support authors throughout the entire storytelling process, from brainstorming and character building to drafting, reviewing, and exporting their work. The platform offers a dedicated “Codex” wiki where writers can catalogue characters, locations, lore, and world‐building elements, ensuring consistency and easy reference. It supports structured planning modes, acts, chapters, scenes, and integrates seamlessly with the writing interface so authors can switch fluidly between planning and prose. AI tools are optional and fully customisable: authors can connect their own API keys (OpenAI, Claude, local LLMs) and set up tailored prompts, or work manually without AI support. Novelcrafter includes a distraction-free write mode, revision history, import/export of documents (Word, Markdown, HTML), and mobile compatibility for writing on the go.

Agent Zero is an open source AI agent framework designed to run autonomous AI assistants that can perform complex tasks by interacting directly with a computer system. It provides an environment where AI agents operate with real system access, allowing them to execute commands, write and run code, browse the web, analyze data, and manage workflows as part of real-world automation processes. Instead of functioning as a simple chat interface, Agent Zero runs in its own virtual environment where it can interact with the operating system, install tools, execute scripts, and coordinate tasks across multiple components. It emphasizes transparency and control, allowing developers to view, modify, and customize how the agent behaves, what tools it can access, and how it processes information. Agent Zero uses a modular architecture that allows the agent to dynamically create and use tools while maintaining persistent memory.

Spacesuit is a productivity and personal thinking tool designed as a “futuristic canvas” where users can organize ideas freely without the constraints of traditional folders or hierarchical systems. It provides a flexible grid-based workspace where anything can be placed anywhere, allowing thoughts, notes, and resources to evolve naturally as users think and create. Instead of rigid structures, it uses resizable “cards” as containers for ideas, which can be moved, resized, and connected across multiple grids to reflect different contexts or workflows. It includes powerful navigation features like a global search bar and grid linking, enabling users to jump between related ideas and access information instantly. Spacesuit integrates AI processors that connect with models from providers such as OpenAI, Anthropic, Google, and others, allowing users to generate insights and dynamic content tailored to their context.

StarCoder and StarCoderBase are Large Language Models for Code (Code LLMs) trained on permissively licensed data from GitHub, including from 80+ programming languages, Git commits, GitHub issues, and Jupyter notebooks. Similar to LLaMA, we trained a ~15B parameter model for 1 trillion tokens. We fine-tuned StarCoderBase model for 35B Python tokens, resulting in a new model that we call StarCoder. We found that StarCoderBase outperforms existing open Code LLMs on popular programming benchmarks and matches or surpasses closed models such as code-cushman-001 from OpenAI (the original Codex model that powered early versions of GitHub Copilot). With a context length of over 8,000 tokens, the StarCoder models can process more input than any other open LLM, enabling a wide range of interesting applications. For example, by prompting the StarCoder models with a series of dialogues, we enabled them to act as a technical assistant.

The next generation of our open source large language model. This release includes model weights and starting code for pretrained and fine-tuned Llama language models — ranging from 7B to 70B parameters. Llama 2 pretrained models are trained on 2 trillion tokens, and have double the context length than Llama 1. Its fine-tuned models have been trained on over 1 million human annotations. Llama 2 outperforms other open source language models on many external benchmarks, including reasoning, coding, proficiency, and knowledge tests. Llama 2 was pretrained on publicly available online data sources. The fine-tuned model, Llama-2-chat, leverages publicly available instruction datasets and over 1 million human annotations. We have a broad range of supporters around the world who believe in our open approach to today’s AI — companies that have given early feedback and are excited to build with Llama 2.

The leading open-source AI code assistant. You can connect any models and any context to create custom autocomplete and chat experiences inside the IDE Remain in flow while coding by removing the barriers that block productivity when building software. Accelerate development with a plug-and-play system that makes it easy to get started and integrates with your entire stack. Become a leader in AI by setting up your code assistant to evolve as new capabilities emerge. Continue autocompletes single lines or entire sections of code in any programming language as you type. Attach code or other context to ask questions about functions, files, the entire codebase, and more. Highlight code sections and press a keyboard shortcut to rewrite code from natural language.