AI / LLMs: The Big List
Date: 2025-08-11
Contents
Introduction
As I'm trying to dig into the huge number of LLM-related tools, I've decided to compile a big list to reference. I'm sharing that list here.
Agentic Tools
Aider
- URL: https://github.com/Aider-AI/aider
- GitHub stars: ~36.5k [GitHub]
- Key feature: AI pair-programming in your terminal.
- Supported LLM Models: Model-agnostic.
- Interaction: Command line.
- Best used for: Local Git-based editing with LLM collaboration.
Amazon Q CLI
- URL: https://github.com/aws/amazon-q-developer-cli
- GitHub stars: ~1.4k
- Key feature: Automated code reviews and transformations via GitHub integration.
- Supported LLM Models: Amazon Q Developer models.
- Interaction: Command line.
- Best used for: GitHub workflows to review or update code automatically.
Claude Code
- URL: https://github.com/anthropics/claude-code
- GitHub stars: ~29.1k
- Key feature: Specialized AI coding agent by Anthropic, rapid adoption (100k users in first week). The Verge
- Supported LLM Models: Anthropic's Claude models.
- Interaction: Command line.
- Best used for: Command-line AI coding with Anthropic-quality outputs.
Cline
- URL: https://cline.bot/ and https://github.com/cline/cline
- GitHub stars: ~49k GitHubCline
- Key feature: Lightweight conversational file-level editing integrated in IDE; "Yolo Mode" bypasses permission dialogs.
- Supported LLM Models: Model-agnostic.
- Interaction: VSCode, Visual Studio, Cursor, and Windsurf extensions.
- Best used for: Focused multi-turn coding edits within your IDE.
Continue
- URL: https://www.continue.dev/ and https://github.com/continuedev/continue
- GitHub stars: ~28.3k
- Key feature: Build custom AI assistants in IDE with extensions and model/rule control. ContinueGitHub
- Supported LLM Models: Model-agnostic, supports any model.
- Interaction: VS Code and JetBrains IDE extensions.
- Best used for: Customizable AI coding agents tailored to your workflow.
Continue CLI
- URL: Part of Continue project; see above.
- GitHub stars: ~28.3k
- Key feature: Modular, battle-tested agent loop for CLI, plug in models/rules/tools. [Continue Docs](https://docs.continue.dev/guides/cli
- Supported LLM Models: Model-agnostic, allows plugging in models.
- Interaction: Command line.
- Best used for: Command-line based custom AI coding agent development.
Crush
- URL: https://github.com/charmbracelet/crush
- GitHub stars: ~8.7k GitHub
- Key feature: Glamorous AI agent for terminal, multi-model and session-based.
- Supported LLM Models: Model-agnostic, supports Anthropic, OpenAI, Gemini, Groq, Claude via AWS Bedrock, OpenAI models hosted on Azure, Llama 3.1, Mistral, Gemma 2.
- Interaction: Command line.
- Best used for: Interactive, fun terminal coding with minimal friction, supporting a wide range of LLMs.
Cursor
- URL: https://www.cursor.com/
- GitHub stars: ~31k
- License: Proprietary
- Price: Free (Hobby plan with limited features), Pro ($20/month or $16/month annually), Ultra ($200/month), Teams ($40/user/month), Enterprise (custom pricing).
- Key feature: An AI-powered code editor built as a fork of VS Code, offering deep AI integration for code generation. Auto mode selects the best model based on the use case, current demand, and output performance.
- Supported LLM Models: Supports frontier coding models including OpenAI (GPT-4o, o3-pro, GPT-4.1, GPT-5), Anthropic (Claude Sonnet 4, Claude Opus 4.1, Claude 3.5 Sonnet, Haiku), Grok, o3, 04-mini and Google (Gemini 2.5 Pro). It also allows for custom model integration via OpenAI API compatible endpoints.
- Interaction: Primarily an integrated development environment (IDE) with AI features accessed via a chat interface, inline edits, and keyboard shortcuts; also offers a command-line interface (CLI).
- Best used for: Boosting developer productivity by providing intelligent code suggestions, automating code writing and refactoring, answering codebase questions, and facilitating debugging within a familiar editor environment.
Gemini CLI
- URL: https://github.com/google-gemini/gemini-cli
- GitHub stars: ~68.5k
- Key feature: AI agent integrated into GitHub Actions for issue triage, PR reviews, and commands; brings Gemini's agentic agent capabilities directly to your terminal.
- Supported LLM Models: Google Gemini models, including Gemini 2.5 Pro.
- Interaction: Command line (terminal), and integrates with GitHub Actions and Gemini Code Assist in VS Code.
- Best used for: Automating GitHub workflows, coding, debugging, data analysis, creative text generation, understanding and navigating large codebases, fixing bugs, writing and testing code, and generating documentation via command line.
Github CoPilot
- URL: https://github.com/features/copilot/
- GitHub stars: N/A
- Price: $10 / month
- Key feature: An AI pair programmer that provides real-time code suggestions, auto-completion, and code generation directly within your integrated development environment (IDE).
- Supported LLM Models: OpenAI Codex models, fine-tuned by GitHub. Specific models like GPT-4 are used for features such as GitHub Copilot Chat.
- Interaction: Primarily via IDE extensions (VS Code, JetBrains IDEs, Visual Studio, Neovim) and also through a chat interface within the IDE.
- Best used for: Accelerating coding by generating code, suggesting completions, writing tests, debugging, and explaining code in various programming languages directly within the editor.
GPTMe
- URL: https://github.com/gptme/gptme
- GitHub stars: ~3.9k
- Key feature: General-purpose agent for all kinds of knowledge-work.
- Supported LLM Models: Model-agnostic.
- Interaction: Command line.
- Best used for: General knowledge work and various automation tasks beyond just coding.
OpenAI Codex
- URL: https://github.com/openai/codex
- GitHub stars: ~34k
- Key feature: An autonomous AI software engineering agent that performs tasks like writing features, answering codebase questions, running tests, and proposing PRs. Also has a local agent for terminal use.
- Supported LLM Models: OpenAI o3, GPT-5 (via ChatGPT Pro, Enterprise, Team, Plus plans), and can use OpenAI API keys.
- Interaction: Command line interface for the local agent, also available via ChatGPT.
- Best used for: Translating natural language to code, refactoring/explaining existing code, generating tests, providing real-time assistance, and automating various software development tasks.
Open Hands
- URL: https://github.com/All-Hands-AI/OpenHands
- GitHub stars: ~61.9k
- Key feature: An open-source AI agent that automates software development tasks.
- Supported LLM Models: Model-agnostic.
- Interaction: Web-based (runs locally).
- Best used for: Automating a wide range of software development tasks.
Plandex
- URL: https://github.com/plandex-ai/plandex
- GitHub stars: ~14.3k
- Key feature: Plans and executes large coding tasks by combining the best models.
- Supported LLM Models: Model-agnostic, supports Anthropic, OpenAI, Google, and open-source providers.
- Interaction: Command line.
- Best used for: Orchestrating and executing complex, multi-model coding projects.
Refact CLI
- URL: https://github.com/smallcloudai/refact
- GitHub stars: ~3.2k
- Key feature: Automates complex, multi-step tasks.
- Supported LLM Models: Model-agnostic; supports OpenAI models (GPT-4, GPT-3.5), Code Llama, and custom fine-tuned models.
- Interaction: Command line interface.
- Best used for: Automating complex, multi-step tasks in coding workflows.
Roo Code
- URL: https://roocode.com/ and https://github.com/RooCodeInc/Roo-Code
- GitHub stars: ~18.4k
- Key feature: Autonomous AI coding agent in editor; offers natural language interaction, file operations, terminal access, browser integration, and customizable personality.
- Supported LLM Models: Model-agnostic; configurable with various models via OpenAI Compatible API, OpenRouter, and Google AI Studio, including GPT-3.5, GPT-4, OpenAI o3, Gemini 2.5 Flash/Pro, DeepSeek models, Mistral Devstral Small, Qwen, Abab, Moonshot, Perplexity, Sonar, Grok-4, Zhipu, Imagen 4 Ultra, Seedance.
- Interaction: VSCode fork IDE.
- Best used for: Editor-embedded multi-agent coding workflows with customizable LLM backends.
Tabby
- URL: https://www.tabbyml.com/ and https://github.com/TabbyML/tabby
- GitHub stars: ~31.9k
- Key feature: Self-hosted AI coding assistant, open-source alternative to Copilot; offers auto-completion, code explanations, and natural language interactions.
- Supported LLM Models: Model-agnostic; compatible with major Coding LLMs (CodeLlama, StarCoder, CodeGen), and can integrate with Jan (local server) and other self-hosted LLMs.
- Interaction: IDE extensions (VSCode, JetBrains, IntelliJ, Android Studio, Vim) and web console.
- Best used for: On-premise, privacy-conscious AI coding assistance, with customizable model backends.
Void
- URL: https://voideditor.com/ and https://github.com/voideditor/void
- GitHub stars: ~26.5k
- Key feature: Open-source Cursor alternative; offers local model support, checkpointing, and privacy-focused features.
- Supported LLM Models: Model-agnostic.
- Interaction: VSCode-based editor.
- Best used for: Developers needing full data control and AI code assistance locally.
Warp
- URL: https://www.warp.dev/ and https://github.com/warpdotdev
- GitHub stars: ~24.3k
- License: Closed source
- Price: Starts at $15 / month
- Key feature: A modern terminal designed for developers that integrates AI capabilities, team collaboration, and a block-based interface, offering AI-powered command suggestions, workflow guidance, and code generation.
- Supported LLM Models: Claude Opus 4.1, Claude Sonnet 4, OpenAI o3, and Gemini 2.5 Pro, and others.
- Interaction: Command line (terminal), featuring an IDE-like text editor, a command palette, and compatibility with various shells like Zsh, Bash, Fish, PowerShell, WSL, and Git Bash.
- Best used for: Boosting productivity for developers and data professionals, enhancing team collaboration through shared sessions and reusable workflows, and streamlining terminal tasks with integrated AI assistance for coding, debugging, and efficient command management.
Zed
- URL: https://zed.dev
- GitHub stars: ~63.3k
- Key feature: Native, high-performance editor designed for AI and collaboration; not a VSCode fork. Includes agentic editing, real-time collaboration with shared cursors, chat, and collaborative debugging.
- Supported LLM Models: Model-agnostic; supports Zed's own hosted models, and API keys for Amazon Bedrock, Anthropic, DeepSeek, GitHub Copilot Chat, Google AI, LM Studio, Mistral, Ollama, OpenAI (including OpenAI API Compatible for other hosted/local models like Together AI, Anyscale), Vercel, xAI (Grok models).
- Interaction: IDE.
- Best used for: High-performance and collaborative coding with integrated AI assistance, especially for users prioritizing speed and local LLM integration.