AI programming tools

Sourcery is an AI-powered code review and refactoring platform designed to improve development efficiency and code quality. It supports over 30 programming languages and integrates with GitHub, GitLab, and major IDEs like VS Code, JetBrains, etc., providing instant code feedback and optimization suggestions. Sourcery automatically identifies potential bugs and security issues, generating PR summaries, review guides, and visual charts to help teams quickly understand code changes. Its built-in Coding Assistant supports generating test cases, documents, code explanations, and real-time refactoring suggestions. The platform supports custom review rules, learns from team feedback, and continuously optimizes review strategies. Sourcery offers strict data security guarantees, supports zero-code retention, and self-hosted deployments, making it suitable for both individual developers and enterprise teams.

Trae is an adaptive AI programming platform launched by ByteDance that aims to improve development efficiency through human-machine collaboration. The platform integrates large language models like Claude 3.7 Sonnet and GPT-4o, supports multilingual development (such as Go, Java, Python, Node.js), and provides zero-configuration environment setups. Trae's core features include intelligent code completion, natural language programming, automatic task dismantling, and multimodal interaction, supporting image uploads to aid in design implementation. Its unique Agent system allows users to create and share custom agents, combined with the MCP protocol to achieve multi-agent collaboration and adapt to different development needs. Trae also offers real-time preview, version control, and one-click deployment capabilities, making it suitable for various scenarios such as front-end, full-stack development, and product design. The platform is currently free and available for macOS and Windows, with a Linux version coming soon.

DeepCode AI is an AI-powered code security analysis platform launched by Snyk, focusing on static application security testing (SAST) and automated code review. The platform combines symbolic AI and generative AI, supports 19+ programming languages, including Java, Python, JavaScript, TypeScript, C/C++, and more, and can identify and automatically fix security vulnerabilities in real-time in developers' IDEs. Its core feature, DeepCode AI Fix, offers up to 80% accuracy in auto-fixes, significantly reducing mean time to fix (MTTR). DeepCode AI ensures data privacy through self-hosted deployment, making it suitable for enterprise-grade security needs. The platform has been adopted by organizations around the world to help developers secure their applications while maintaining efficient coding.

Amazon CodeWhisperer is an AI programming assistant from AWS designed to enhance developer productivity with real-time code suggestions. It supports a wide range of programming languages, including Python, Java, JavaScript, TypeScript, C#, Go, and more, and is compatible with mainstream IDEs such as Visual Studio Code, JetBrains Series, AWS Cloud9, SageMaker Studio, and more. CodeWhisperer generates code suggestions for entire lines or full functions based on context, assisting in creating AWS resources (e.g., S3, DynamoDB), writing unit tests, importing libraries, and managing AWS sessions. It also has security scanning capabilities that identify common vulnerabilities such as the OWASP Top 10 and provide remediation recommendations. CodeWhisperer offers a free Personal and Pro version, with the Pro version supporting team management, customization, and enterprise-grade security controls. CodeWhisperer's capabilities are now being integrated into Amazon Q Developer, which provides a more powerful generative AI programming experience.

Bolt.new is an AI-powered, full-stack web development platform from StackBlitz that allows users to quickly build, edit, run, and deploy full web or mobile apps in their browsers with just natural language prompts. The platform is based on WebContainers technology and supports mainstream frameworks such as Node.js, React, Next.js, Vue, and Tailwind, eliminating the need for local environment configuration. Bolt.new integrates Anthropic's Claude model, which understands user intent, automatically generates front-end and back-end code structures, and provides real-time previews and one-click deployment capabilities. Additionally, the platform supports integration with tools like Figma, Supabase, Stripe, GitHub, and more, making it suitable for rapid prototyping, instructional demonstrations, and small to medium-sized project development. Bolt.new offers both free and paid subscription options, catering to different user development needs.

Golex AI is a no-code AI website generation platform that allows users to create fully functional web applications in minutes with just natural language descriptions. The platform supports building various types of websites such as e-commerce platforms, blogs, portfolios, etc., suitable for startups, product managers, and non-technical users. Golex AI offers multilingual support, built-in authentication, database management, file storage, payment processing, and other functions, and can be integrated with services such as Firebase and Google Vertex AI to enhance the intelligence of applications. Users can customize their domain names and quickly deploy them online, and the platform offers a variety of subscription plans such as Free, Basic ($20/month), Professional ($50/month), and Enterprise ($100/month) to meet the needs of projects of different sizes.

v0 is an AI-powered front-end development platform launched by Vercel designed to quickly generate high-quality React UI components and full pages from natural language prompts. Users only need to describe the required interface, and v0 can generate code based on Next.js, Tailwind CSS, and shadcn/ui, supporting real-time preview and one-click deployment to Vercel. The platform also supports importing designs from Figma or screenshots, automatically converting them into editable code structures for rapid prototyping and product iteration. v0 offers version control, project management, and multiple integration options, making it suitable for developers, designers, and product managers to work together. Its pricing includes free and multiple paid plans, catering to the development needs of different users.

MGX (MetaGPT X) is an AI programming platform based on a multi-agent architecture designed to automate the entire process from requirements analysis to deployment by simulating real software development processes. The platform introduces multiple AI roles such as product managers, architects, engineers, and data analysts to collaboratively complete the construction of websites, blogs, e-commerce, data analysis, games, and other applications. Users only need to describe their requirements through natural language, and MGX can generate a complete code structure and deployment plan, supporting real-time preview and one-click deployment. The platform also offers a visual editor and Dev Mode for easy interface tweaks and code editing. MGX integrates large language models like GPT-4, Claude, and Gemini, supporting multiple programming languages and frameworks, making it suitable for individual developers, startups, and product teams.

Lovable.dev is an AI-powered full-stack app building platform that allows users to quickly generate complete web applications or websites with just natural language descriptions. The platform integrates large language models such as GPT-4, Claude, and Gemini, supporting automatic generation and deployment of front-end (React, Tailwind CSS), back-end, database (Supabase), and API. Users can adjust the interface through the visual editor or edit the code directly in Dev Mode for a high degree of customization. Lovable.dev offers features like real-time collaboration, version control, GitHub integration, Figma import, custom domains, and more, making it suitable for individual developers, startups, and product teams. Its free plan offers 5 generations per day, with paid plans starting at $25/month, catering to different user development needs.

HaiSnap is an AI zero-code application generation platform launched by Beijing HaiSnap Intelligent Technology Co., Ltd., dedicated to enabling everyone to create their own digital tools, web pages, or games in natural language. Users do not need programming experience, just describe their requirements in one sentence, and the platform can automatically generate a complete web application, and supports one-click deployment and source code download, which is convenient for users to carry out secondary development or learning. HaiSnap provides visual operations and a drag-and-drop interface that supports calling a variety of pre-trained AI models, covering functions such as text generation and image recognition. The platform also integrates modules such as task management, gadgets and mini-games, which are suitable for various scenarios such as education, entertainment, and internal tool development to help users efficiently complete tasks and improve work efficiency.

LlamaCoder is an open-source AI programming platform launched by Together AI, designed to quickly generate complete web applications from natural language prompts. The platform is built on Meta's Llama 3.1 405B model, which supports the generation of React applications and components, and users only need to describe the application requirements, and LlamaCoder can generate corresponding code. The platform offers both online and on-premises deployments, catering to the needs of different developers. LlamaCoder integrates technology stacks such as Sandpack, Next.js, Tailwind CSS, etc., supporting real-time preview and one-click deployment, suitable for rapid prototyping and product iteration. Its open-source nature and active community support make it an ideal tool for developers exploring AI programming.

CircleCI is a leading continuous integration and continuous delivery (CI/CD) platform designed to help development teams automate the build, test, and deployment process to accelerate software delivery cycles. The platform supports a variety of programming languages and frameworks, including Go, Java, Python, Node.js, etc., and is compatible with mainstream version control systems such as GitHub, Bitbucket, and GitLab. Users can define pipelines through configuration files (config.yml) and use Orbs to achieve configuration reuse and improve development efficiency. CircleCI offers both cloud and on-premises deployment options, supporting various execution environments such as Docker, Linux, macOS, Windows, and more, catering to teams of different sizes. Its Insights Dashboard provides real-time monitoring and analytics to help teams optimize their build processes and improve code quality. As a key tool in DevOps practices, CircleCI has been adopted by many enterprises, including Meta, Adobe, Spotify, and more, to enable efficient and reliable software delivery.

Tabnine is a leading AI programming assistant designed to enhance development efficiency and code quality through intelligent code completion and natural language interaction. It supports over 80 programming languages and mainstream IDEs, including VS Code, JetBrains Family Bucket, Visual Studio, Eclipse, and more, making it suitable for a wide range of development scenarios. Tabnine offers multiple deployment methods, including SaaS, VPC, and on-premises, ensuring code privacy and security. Its AI models are trained solely on open-source licensed code, avoiding copyright risks. Tabnine's core features include code generation, auto-completion, bug fixing, unit test generation, documentation writing, and code interpretation, supporting a variety of large language models such as Claude 3.5 Sonnet, GPT-4o, Command R+, and Codestral. Additionally, Tabnine offers a wide range of AI agents that cover the entire software development lifecycle, helping development teams achieve efficient and reliable software delivery.

Manus AI is a startup developed by Singaporean startup Butterfly Effect PTE. LTD. is a general-purpose AI agent platform designed to turn user ideas into actionable tasks. Users only need to describe their needs through natural language, and Manus can independently complete the entire process from planning, research, to execution, covering web development, data analysis, content creation, educational assistance and other fields. The platform supports multi-modal input, with browser control, code generation, terminal operation, and other capabilities, making it suitable for individual users and enterprise teams. Manus offers a free trial with 300 credits per day and a one-time 1,000 credits, accessible to users across multiple platforms including web, iOS, and Android. Its powerful task execution capabilities and flexible interaction methods make it an ideal tool for improving work efficiency and automating operations.

YouWare is an AI programming platform for creators designed to help users quickly build, deploy, and share web applications through natural language prompts and AI tools. The platform supports uploading or pasting code to automatically generate shareable web links, enabling seamless conversion from idea to presentation. Users can also use AI Coding Agent to automate page design, coding, and deployment by describing requirements, lowering the barrier to entry for non-technical users. YouWare offers one-click beautification to optimize web design and make it more professional. The platform encourages creative sharing and supports the project's Remix feature, fostering community collaboration and innovation. Additionally, YouWare introduces a "Knots" points system that rewards creators based on project visits, interactions, and remixes, with points redeemable for cash and incentivizing the generation of high-quality content. The platform also provides privacy controls, supporting public, private, or password-protected sharing of projects to meet the needs of different users. YouWare is committed to creating an active, supportive creative community where everyone can easily bring their ideas to life.

Devv AI is an AI search engine and coding assistant designed for developers, designed to provide fast and accurate programming solutions through advanced language models such as GPT-4, Claude 3.5, Gemini. The platform supports multiple search modes, including Chat Mode (direct conversation with AI), Web Mode (provides web-based answers), and GitHub Mode (integrates with GitHub repositories for contextual support). Additionally, Devv AI offers the VS Code plugin, allowing users to access AI-assisted coding support directly within their development environment. The platform also launched Devv Builder, which automatically generates complete back-end code, database configuration, and deployment scenarios by simply describing requirements, suitable for rapid prototyping and application development. Devv AI offers both free and paid subscription options, catering to different user development needs.

Cline is an open-source AI programming assistant designed for Visual Studio Code, featuring features such as Plan/Act mode, terminal command execution, and Model Context Protocol (MCP) integration. It enables deep understanding of the entire codebase, multi-file editing, running tests, debugging browsers, and seamless integration with project management tools. Cline supports a wide range of popular AI models, including Claude 3.7 Sonnet, DeepSeek Chat, and Google Gemini 2.0 Flash, allowing users to choose the right model based on their needs. With MCP, Cline can be scaled to connect to external databases, APIs, and documentation, enabling highly automated development processes. Its permission control mechanism ensures that every step of the operation requires user approval, ensuring code security. Cline is free to use, and users only pay for API usage for their chosen model, making it suitable for both individual developers and enterprise teams.

Inception Labs is an AI startup founded by professors from Stanford, UCLA, and Cornell University that focuses on developing next-generation large language models (dLLMs) based on diffusion models. Its core product, the Mercury series of models, adopts a parallel generation mechanism to significantly improve the speed and efficiency of text and code generation, with a single-user throughput of up to 1,000 tokens per second, outperforming mainstream models such as GPT-4o and Claude 3.5 Haiku. Inception Labs provides OpenAI-compatible APIs that support local deployment and model fine-tuning, and has been integrated with platforms such as Microsoft NLWeb and Coplay, and is widely used in scenarios such as code completion, enterprise automation, and natural language interfaces. The company is committed to promoting the widespread application of generative AI in multimodal fields through faster and lower-cost AI solutions.

Replicate is an AI model hosting and inference platform for developers, providing a unified API interface that supports various generation tasks such as images, videos, audio, and text. Users can easily call open source models on the platform, including Stable Diffusion, FLUX, Whisper, GPT-4, etc., through client libraries such as Python and Node.js, to achieve functions such as text generation to images, speech-to-text, video generation, and text generation. Replicate also supports model fine-tuning and custom deployment, suitable for AI drawing, AI video, AI audio, AI writing, AI programming, AI design, AI detection, and other fields. The platform offers a predictable billing model, making it suitable for both individual developers and business users.

Roo Code is an open-source AI programming assistant that is integrated into the development environment as a VS Code extension, providing a multi-agent collaboration experience for development teams. It supports natural language interaction, with the ability to read and edit project files, execute terminal commands, automate browser actions, and integrate custom tools through the Model Context Protocol (MCP). Roo Code has a variety of built-in dedicated modes, such as Code, Architect, Debug, Ask, etc., and users can also create custom patterns to meet specific development needs. Its Intelligent Context Condensing function can automatically manage the contextual information of large projects and improve processing efficiency. The platform supports a variety of AI models, including OpenAI, Google Gemini, Anthropic Claude, etc., and can run local models through LM Studio, suitable for scenarios with high privacy and performance requirements. Roo Code is widely used in code generation, refactoring, debugging, documentation, and other tasks, making it suitable for developers, AI researchers, and automation engineers.

Warp Terminal is a modern terminal built in Rust language, integrating the AI assistant "Agent Mode" to support natural language instructions, code generation, error interpretation, and multi-step task execution. Its IDE-style editor offers multi-line editing, intelligent completion, and command block management to improve development efficiency. The Warp Drive feature allows users to save and share commands, workflows, and environment variables, facilitating team collaboration. Compatible with zsh, bash, fish, and PowerShell, supporting macOS, Linux, and Windows platforms, it is suitable for various scenarios such as developers, DevOps engineers, and AI researchers.

Firecrawl is an AI-powered web scraping and data extraction platform for developers that transforms any website content into structured data for large language models (LLMs). The platform offers a variety of features, including single-page scraping (/scrape), full-site crawling (/crawl), website structure mapping (/map), web search (/search), and structured data extraction (/extract), supporting various output formats such as Markdown, JSON, and HTML. Firecrawl has the ability to handle dynamic JavaScript content, bypass anticrawling mechanisms, and automatic proxy rotation, making it suitable for building RAG systems, AI assistants, knowledge bases, and market intelligence tools. The platform provides a variety of SDKs such as Python, Node, Go, and Rust, and is deeply integrated with AI frameworks such as LangChain, LlamaIndex, and Dify, supporting local deployment and open source hosting, suitable for a variety of application scenarios from individual developers to enterprise-level users.

Portkey AI is a full-stack platform for generative AI applications, providing a unified AI gateway, observability, prompt management, daemon mechanism, and MCP client to help developers efficiently build and deploy LLM applications. The platform supports connecting more than 250 models and has functions such as load balancing, conditional routing, automatic retries, and semantic caching to improve system stability and performance. Portkey AI also offers real-time monitoring and log tracking, helping teams optimize costs and response quality. Its open-source architecture supports on-premises deployment, making it suitable for enterprise users with high requirements for data security and compliance.

Modal is a high-performance AI infrastructure-as-a-service platform that provides a single line of Python code to run functions in the cloud, enabling serverless automatic scaling for machine learning inference, model fine-tuning, and large-scale batch processing. The platform supports custom container images, one-click scheduling of GPU instances such as A100, H100, and B200, second-level cold start and elastic computing power management, and integrates cloud storage mounting, log monitoring, and HTTPS endpoints. Fine billing is based on CPU/GPU usage time, and $30 free computing power is given per month, without the need to write configuration files, helping developers quickly launch AI applications and optimize costs and efficiency.

Mem0 is a self-optimizing memory layer for developers, providing persistent contextual storage and retrieval services for various large language model applications. Through vectorized retrieval and intelligent data filtering, Mem0 helps scenarios such as customer service bots, personal assistants, AI agents, and e-commerce recommendation systems reduce call costs by up to 80% while retaining critical historical information. The platform offers both fully managed and open-source managed modes, compatible with mainstream models such as OpenAI and Anthropic Claude, and can be quickly integrated and continuously iterated with just a few lines of code, making AI applications more personalized, coherent, and efficient.

Baidu Comate is Baidu's intelligent programming assistant based on the Wenxin model, providing automatic code generation, unit testing, comment generation, intelligent Q&A and security detection functions, and supports 100+ programming languages and multiple IDE integrations. The first one-click conversion of design drafts, cross-file collaboration, and knowledge base enhancement greatly improve programming efficiency and lower the development threshold, helping enterprises and developers quickly deliver high-quality software projects.

Tongyi Lingma is an intelligent coding assistant launched by Alibaba Cloud, which provides efficient automatic code generation and intelligent Q&A capabilities based on the Tongyi model. The platform supports multi-language and multi-framework, with cross-file semantic understanding, auto-completion functions, batch refactoring, and test case generation. Seamlessly integrate VS Code, JetBrains series, and Web IDEs to synchronize code and comments in real time for multi-person collaboration. At the same time, it provides access to privatized knowledge bases and permission management to ensure enterprise source code security and compliance. Tongyi Lingma helps the development team significantly improve R&D efficiency, reduce repetitive labor, and achieve a full-process intelligent programming experience from demand to delivery.

iFlyCode is an intelligent programming assistant independently developed by iFLYTEK, based on iFLYTEK Spark cognitive model, covering the whole process of requirements analysis, design, coding and self-testing, and providing code generation, continuation, completion, error correction, interpretation, function annotation and unit testing. Compatible with mainstream languages such as Java, Python, C++, JavaScript, etc., and deeply integrated with VS Code, JetBrains series IDEs and Web IDEs; It supports SaaS public cloud and private deployment, and has cross-file context understanding and natural language question answering capabilities, significantly improving development efficiency and code quality.

aiXcoder is an intelligent software development tool launched by Sixin Technology, relying on deep learning and large-scale code training, providing functions such as natural language to method-level code generation, token-level and row-level intelligent completion, and cross-file intelligent search. It supports Java, Python, C++, JavaScript, Go and other languages, and is compatible with VS Code, IntelliJ IDEA, PyCharm, Android Studio and other mainstream IDEs, and can complete code snippet generation, unit testing, and automatic comment addition without leaving the editor. Privatization ensures enterprise code security and helps teams significantly improve development efficiency and code quality.

Raccoon family Raccoon is an AI intelligent productivity series launched by SenseTime, based on SenseTime's self-developed large language model, covering two major members: "Code Raccoon" and "Office Raccoon". Code Raccoon supports requirements analysis, architecture design, code writing and testing, and is compatible with 100+ mainstream languages such as Python, Java, C++, as well as VS Code and IDEA plug-ins, helping developers speed up by more than 50%. The office raccoon can connect multiple files and complex tables to achieve data cleaning, calculation, trend analysis and visualization, and significantly improve the efficiency of data processing and reporting. This series has served hundreds of thousands of enterprise users and is widely used in software research and development and intelligent office scenarios.

Chef is a full-stack AI application generator launched by Convex, allowing developers to automatically complete data model design, back-end functions, real-time databases, authentication systems, and front-end interface construction by entering a single request, and preview it in real time in the browser. The platform has built-in components such as file storage, scheduled tasks, and multi-person collaboration, which can be deployed to the cloud with one click, and also supports exporting TypeScript code to connect with GitHub for continuous integration to ensure production-level stability and maintainability. Chef is suitable for startup teams, hackathons, and enterprises to quickly validate SaaS, social, gaming, and AI agents, significantly reducing development thresholds and costs, and helping developers iterate and launch high-quality applications in minutes.