[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-83967":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":9,"languages":9,"totalLinesOfCode":9,"stars":10,"forks":11,"watchers":12,"openIssues":13,"contributorsCount":13,"subscribersCount":13,"size":13,"stars1d":14,"stars7d":15,"stars30d":15,"stars90d":13,"forks30d":13,"starsTrendScore":16,"compositeScore":17,"rankGlobal":9,"rankLanguage":9,"license":18,"archived":19,"fork":19,"defaultBranch":20,"hasWiki":21,"hasPages":19,"topics":22,"createdAt":9,"pushedAt":9,"updatedAt":23,"readmeContent":24,"aiSummary":9,"trendingCount":13,"starSnapshotCount":13,"syncStatus":12,"lastSyncTime":25,"discoverSource":26},83967,"awesome-free-models","12britz\u002Fawesome-free-models","12britz","A curated list of free AI models, APIs, and tools you can use without paying a cent.",null,182,13,2,0,11,52,74,88.54,"Other",false,"main",true,[],"2026-06-12 04:01:42","# Awesome Free Models [![Awesome](https:\u002F\u002Fawesome.re\u002Fbadge-flat.svg)](https:\u002F\u002Fawesome.re)\n\n> A curated list of free AI models, APIs, and tools you can use without paying a cent.\n\n![Last Updated](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLast%20Checked-June%208%2C%202026-brightgreen?style=for-the-badge)\n![Models](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FModels-29-blue?style=flat-square)\n![Tools](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FTools-150-blue?style=flat-square)\n![Sections](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSections-15-blue?style=flat-square)\n![License](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-CC0-lightgrey?style=flat-square)\n\n> ✅ All links verified live on June 8, 2026. 6 broken links found and fixed during this check.\n\nRunning AI shouldn't require a credit card. This list curates genuinely free models — open-weight models you can self-host, free API tiers from major providers, and tools to run everything locally.\n\n---\n\n## Contents\n\n- [🧠 Open-Weight Models](#-open-weight-models)\n- [🔌 Free API Providers](#-free-api-providers)\n- [💻 Local Inference Tools](#-local-inference-tools)\n- [💬 AI Chatbot UIs](#-ai-chatbot-uis)\n- [🤖 AI Coding Assistants](#-ai-coding-assistants)\n- [📝 Code Models](#-code-models)\n- [🔍 RAG & Vector Databases](#-rag--vector-databases)\n- [🧩 Agentic Frameworks](#-agentic-frameworks)\n- [🎛 Fine-tuning Tools](#-fine-tuning-tools)\n- [✨ Prompt Engineering Tools](#-prompt-engineering-tools)\n- [📊 Datasets](#-datasets)\n- [☁ Model Hosting Platforms](#-model-hosting-platforms)\n- [📚 Learning Resources](#-learning-resources)\n- [🏆 Resources & Leaderboards](#-resources--leaderboards)\n- [👥 Communities](#-communities)\n\n---\n\n## 🧠 Open-Weight Models\n\n> 📅 Last checked: June 8, 2026\n\nNotable open-weight models you can download and run on your own hardware.\n\n| Name | Description |\n|------|-------------|\n| [Llama 4 Scout \u002F Maverick](https:\u002F\u002Fhuggingface.co\u002Fmeta-llama) | Meta's latest MoE generation. Scout: 109B, 10M context. Maverick: 402B, 1M context. Native multimodal. [[License]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fllama\u002Fblob\u002Fmain\u002FLICENSE) |\n| [DeepSeek V4](https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai) | Latest generation with extreme cost-efficiency. MIT license. |\n| [DeepSeek-V4-Flash](https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai\u002FDeepSeek-V4-Flash) | **Apr 2026.** Efficiency-focused variant of DeepSeek V4. 1M token context, optimized for fast inference. MIT license. |\n| [Gemma 4 31B \u002F 26B MoE \u002F E4B \u002F E2B](https:\u002F\u002Fhuggingface.co\u002Fgoogle) | Fully permissive Apache 2.0. 256K context, native multimodal. New standard for open-weight. |\n| [GLM-5.1 (Zhipu AI)](https:\u002F\u002Fhuggingface.co\u002FTHUDM) | 744B MoE model, competitive with top proprietary models. MIT license. |\n| [MiniMax M3](https:\u002F\u002Fhuggingface.co\u002FMinimax) | Frontier-tier 1M context, native multimodal + computer use. MSA architecture. |\n| [Trinity (Arcee AI)](https:\u002F\u002Fhuggingface.co\u002Farcee-ai) | 400B parameter enterprise model. Apache 2.0. |\n| [Step 3.7 Flash (StepFun)](https:\u002F\u002Fhuggingface.co\u002Fstepfun-ai) | **May 2026.** Apache 2.0. Native multimodal (image+video), strong agentic performance. Efficient enough for high-end local hardware. |\n| [Kimi K2.6 (Moonshot AI)](https:\u002F\u002Fhuggingface.co\u002Fmoonshotai) | **Apr 2026.** 1T-parameter MoE model. Modified MIT license. Exceptional coding (SWE-Bench ~54%) and multi-agent swarm orchestration. |\n| [Qwen 3.6-35B-A3B](https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen3.6-35B-A3B) | **Apr 2026.** MoE variant with only 3B active parameters. Extremely efficient for consumer hardware. Apache 2.0. |\n| [InternLM 3 (Shanghai AI Lab)](https:\u002F\u002Fhuggingface.co\u002Finternlm) | **Early 2026.** Strong long-context reasoning and agentic performance. Competitive in open-weight benchmarks. |\n| [MiMo-V2.5-Pro (Xiaomi)](https:\u002F\u002Fhuggingface.co\u002FXiaomiMiMo\u002FMiMo-V2.5-Pro) | **Apr 2026.** 1.02T-parameter MoE (42B active). Optimized for complex agentic tasks, coding, and long-context. |\n| [Bonsai 8B (PrismML)](https:\u002F\u002Fhuggingface.co\u002Fprism-ml\u002FBonsai-8B-gguf) | **Apr 2026.** Groundbreaking 1-bit quantized model. Extremely efficient for edge and consumer hardware (Apple Silicon). |\n| [Mistral Small 3.1 (Mistral)](https:\u002F\u002Fhuggingface.co\u002Fmistralai\u002FMistral-Small-3.1-24B-Instruct-2503) | **Mar 2025.** Versatile 24B multimodal model. Strong text performance with native image understanding and 128K context. Apache 2.0. |\n| [Mistral Small 4 (Mistral)](https:\u002F\u002Fhuggingface.co\u002Fmistralai\u002FMistral-Small-4-119B-2603) | **Mar 2026.** Hybrid MoE (6.5B active params) unifying instruction, reasoning, and multimodal capabilities. Efficient frontier-class model. Apache 2.0. |\n| [Command A+ (Cohere)](https:\u002F\u002Fhuggingface.co\u002FCohereLabs\u002Fcommand-a-plus-05-2026-w4a4) | **May 2026.** Enterprise multimodal MoE optimized for sovereignty and multilingual RAG across 48 languages. Apache 2.0. |\n| [Hermes 4 (NousResearch)](https:\u002F\u002Fhuggingface.co\u002FNousResearch\u002FHermes-4-70B) | **Feb 2026.** Self-improving agentic model with closed-loop learning. Curates own memory and builds skills from experience. Apache 2.0. |\n| [Snowflake Arctic](https:\u002F\u002Fhuggingface.co\u002FSnowflake\u002Farctic) | **Apr 2024.** Enterprise MoE model balancing high-quality performance with efficient training costs. Optimized for complex data operations. Apache 2.0. |\n| [Falcon 3 (TII)](https:\u002F\u002Fhuggingface.co\u002Ftiiuae\u002FFalcon3-7B-Instruct) | **Dec 2024.** Compact high-performance model with strong reasoning. Designed for efficient deployment on resource-constrained hardware. TII Falcon-LLM License 2.0. |\n| [Apple OpenELM](https:\u002F\u002Fhuggingface.co\u002FApple\u002FOpenELM-3B) | **Apr 2024.** Family of efficient on-device SLMs using layer-wise attention scaling. Runs locally on Apple Silicon with full privacy. Apple Sample Code License. |\n\n---\n\n## 🔌 Free API Providers\n\n> 📅 Last checked: June 8, 2026\n\nProviders offering free tiers to access models via API — no local hardware required.\n\n| Name | Description |\n|------|-------------|\n| [Google AI Studio](https:\u002F\u002Faistudio.google.com\u002F) | **Most generous free tier.** Access Gemini 2.5 Flash, Gemini 2.0 Flash, and other models. Generous rate limits for prototyping. |\n| [OpenRouter](https:\u002F\u002Fopenrouter.ai\u002F) | Aggregates 500+ models. Filter by \"Free\" to see models available at no cost. Includes experimental and subsidized open-weight models. |\n| [Groq](https:\u002F\u002Fconsole.groq.com\u002F) | Ultra-fast inference. Free tier includes Llama, Gemma, Mixtral, Whisper models with generous daily rate limits. |\n| [Hugging Face Inference API](https:\u002F\u002Fhuggingface.co\u002Finference-api) | Free tier for thousands of community models. Rate-limited but excellent for testing. |\n| [NVIDIA NIM](https:\u002F\u002Fbuild.nvidia.com\u002F) | Free API access to accelerated versions of Llama, Mistral, Gemma, and more on NVIDIA infrastructure. |\n| [DeepInfra](https:\u002F\u002Fdeepinfra.com\u002F) | Serverless inference. Free tier with daily rate limits for popular open-source models. |\n| [Together AI](https:\u002F\u002Fwww.together.ai\u002F) | Free trial credits for new users. Fast inference on open-source models. |\n| [Fireworks AI](https:\u002F\u002Ffireworks.ai\u002F) | Free tier for community models. Optimized for low latency. |\n| [SiliconFlow](https:\u002F\u002Fsiliconflow.cn\u002F) | Rising platform with free access to many open-source models. |\n| [Cloudflare Workers AI](https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers-ai\u002F) | Free tier for running select open-source models at the edge. |\n| [Replicate](https:\u002F\u002Freplicate.com\u002F) | Free tier with limited credits for running open-source models. |\n| [Poe (Quora)](https:\u002F\u002Fpoe.com\u002F) | Free tier with daily credits for GPT-4 mini, Claude instant, and community bots. |\n| [CatGPT](https:\u002F\u002Fwww.catgpt.cc\u002F) | Completely free chat with multiple models, no login required. ⚠️ Currently unreachable (SSL cert expired as of June 2026). |\n| [Qwen Chat (Alibaba)](https:\u002F\u002Fchat.qwen.ai\u002F) | Free access to Qwen 3.6-Plus, Qwen 3.6-Max, and other Qwen models via web chat and API. 1M token context for agentic coding. |\n| [Ollama Cloud](https:\u002F\u002Follama.com\u002Fcloud) | Free tier for running open-source models on Ollama's cloud infrastructure. Light usage included, 1 concurrent model. Same `ollama run` command as local. Zero data retention. |\n| [Mistral AI (La Plateforme)](https:\u002F\u002Fmistral.ai\u002F) | Free API tier with access to Mistral Large, Mistral Nemo, Codestral and more. 1 req\u002Fs, 500k tokens\u002Fmin. Requires phone verification and data usage opt-in. |\n| [Cohere](https:\u002F\u002Fcohere.com\u002F) | Free evaluation API key for Command R, Command R+, Embed, and Rerank models. 20 req\u002Fmin, 1,000 req\u002Fmonth. |\n| [DeepSeek Platform](https:\u002F\u002Fdeepseek.com\u002F) | Free API credits for new users (5M tokens). Access to DeepSeek V4, DeepSeek-R1, and other models. Generous free allocation. |\n| [GitHub Models](https:\u002F\u002Fgithub.com\u002Fmarketplace\u002Fmodels) | Free tier for GitHub users. Access GPT-4o, Llama 3.3, Mistral, and more with rate-limited playground and API. |\n| [Hyperbolic](https:\u002F\u002Fhyperbolic.xyz\u002F) | Open-access AI cloud with affordable inference. Free compute credits via referral program. Supports Llama, Qwen, DeepSeek, and other open models. |\n| [Novita AI](https:\u002F\u002Fnovita.ai\u002F) | Free credits for testing 100+ models including Llama, Qwen, DeepSeek, and Mistral. OpenAI-compatible API with competitive pricing beyond the free tier. |\n| [Anakin.ai](https:\u002F\u002Fanakin.ai\u002F) | **30 daily free credits** for accessing multiple AI models. Web chat interface and API access. Supports GPT-4, Claude, and open-weight models. |\n| [Nebius AI](https:\u002F\u002Fnebius.com\u002F) | **$100 free credits** for new users. AI Studio with access to Llama, Qwen, DeepSeek, and other open-weight models. Fast inference on NVIDIA H100 infrastructure. |\n| [Fal.ai](https:\u002F\u002Ffal.ai\u002F) | Free starter credits for AI inference. Fast, serverless platform supporting Llama, Flux, and Stable Diffusion models. Pay-as-you-go beyond free tier. |\n| [Vercel AI Gateway](https:\u002F\u002Fvercel.com\u002Fai) | **$5\u002Fmonth free credits** for the AI Gateway. Proxy and cache requests across multiple LLM providers. SDK is open-source and free. |\n| [AI21 Labs](https:\u002F\u002Fwww.ai21.com\u002F) | **$10 trial credits** for accessing Jamba 1.5, Jamba 1.6, and other AI21 models. Valid for 3 months. Requires account sign-up. |\n| [Amazon Bedrock](https:\u002F\u002Faws.amazon.com\u002Fbedrock\u002F) | **$200 AWS credits** for new customers. Access to Llama, Mistral, Claude, Titan, and other foundation models via API. |\n| [Azure AI Foundry](https:\u002F\u002Fazure.microsoft.com\u002Fen-us\u002Fproducts\u002Fai-foundry\u002F) | **$200 free trial credits** (30 days). Access GPT-4o, Llama, Mistral, Phi, and other models via Azure's unified AI platform. |\n| [RunPod](https:\u002F\u002Frunpod.io\u002F) | Free credits for serverless GPU inference. Deploy open-weight models as serverless endpoints. Supports Llama, Qwen, DeepSeek, and more. |\n| [OpenCode](https:\u002F\u002Fopencode.ai\u002F) | Go-based terminal AI coding assistant. Model-neutral, supports multiple LLM providers, LSP integration, and MCP tools. Free and open-source. [GitHub](https:\u002F\u002Fgithub.com\u002Fopencode-ai\u002Fopencode) |\n\n---\n\n## 💻 Local Inference Tools\n\n> 📅 Last checked: June 8, 2026\n\nRun models on your own machine — no API keys needed, full privacy.\n\n| Name | Description |\n|------|-------------|\n| [Ollama](https:\u002F\u002Follama.com\u002F) | The easiest way to run local LLMs. One command to download and run any model. macOS, Linux, Windows. [GitHub](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) |\n| [LM Studio](https:\u002F\u002Flmstudio.ai\u002F) | Polished desktop GUI. Browse, download, and chat with models. Built-in model browser and local API server. |\n| [llama.cpp](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp) | High-performance C++ inference engine. Runs on CPU and GPU. Supports GGUF quantization. Powers most other local tools. |\n| [Jan](https:\u002F\u002Fjan.ai\u002F) | Open-source ChatGPT alternative for desktop. Built-in model downloader, local API server. [GitHub](https:\u002F\u002Fgithub.com\u002Fjanhq\u002Fjan) |\n| [GPT4All](https:\u002F\u002Fgpt4all.io\u002F) | Privacy-focused local chatbot. Runs on consumer hardware. Built-in model browser. [GitHub](https:\u002F\u002Fgithub.com\u002Fnomic-ai\u002Fgpt4all) |\n| [text-generation-webui (Oobabooga)](https:\u002F\u002Fgithub.com\u002Foobabooga\u002Ftext-generation-webui) | Feature-rich web UI. Supports multiple backends (Transformers, llama.cpp, ExLlama, AutoGPTQ). |\n| [LocalAI](https:\u002F\u002Flocalai.io\u002F) | Drop-in OpenAI API replacement. Run models locally with an OpenAI-compatible API. [GitHub](https:\u002F\u002Fgithub.com\u002Fmudler\u002FLocalAI) |\n| [KoboldCPP](https:\u002F\u002Fgithub.com\u002FLostRuins\u002Fkoboldcpp) | Single-file executable for running GGUF models. Focused on story generation but general-purpose. |\n| [llamafile (Mozilla)](https:\u002F\u002Fgithub.com\u002FMozilla-Ocho\u002Fllamafile) | Distributable single-file executables that run LLMs. No installation needed. |\n| [vLLM](https:\u002F\u002Fgithub.com\u002Fvllm-project\u002Fvllm) | High-throughput production inference engine. Uses PagedAttention for efficient serving. |\n| [SGLang](https:\u002F\u002Fgithub.com\u002Fsgl-project\u002Fsglang) | Fast inference framework with structured generation and RadixAttention. |\n| [TensorRT-LLM (NVIDIA)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FTensorRT-LLM) | NVIDIA's optimized inference engine. Best performance on NVIDIA GPUs. |\n| [ExLlamaV2](https:\u002F\u002Fgithub.com\u002Fturboderp\u002Fexllamav2) | Optimized inference for Llama-family models. Fastest option for single-GPU inference. |\n| [Aphrodite Engine](https:\u002F\u002Fgithub.com\u002FPygmalionAI\u002Faphrodite-engine) | High-performance LLM serving engine with advanced quantization support. |\n| [TabbyAPI](https:\u002F\u002Fgithub.com\u002Ftheroyallab\u002FtabbyAPI) | Lightweight, fast OpenAI-compatible API server for ExLlamaV2. |\n| [LlamaEdge](https:\u002F\u002Fllamaedge.com\u002F) | Lightweight inference framework for edge devices. OpenAI-compatible API for open-source models. Runs on WasmEdge for portability. [GitHub](https:\u002F\u002Fgithub.com\u002FLlamaEdge\u002FLlamaEdge) |\n| [MLC LLM](https:\u002F\u002Fgithub.com\u002Fmlc-ai\u002Fmlc-llm) | Universal deployment engine by UW\u002FSJTU. Runs LLMs on any hardware — laptops, phones, browsers. OpenAI-compatible API. |\n| [WebLLM](https:\u002F\u002Fgithub.com\u002Fmlc-ai\u002Fweb-llm) | In-browser LLM inference via WebGPU. Runs models directly in your browser with zero setup. No server needed. |\n| [FastChat (LMSYS)](https:\u002F\u002Fgithub.com\u002Flm-sys\u002FFastChat) | Open platform for training, serving, and evaluating LLMs. Provides OpenAI-compatible API and web UI for local models. |\n| [Hugging Face TGI](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftext-generation-inference) | Production-grade serving toolkit for large language models. Optimized for high throughput on local hardware. |\n| [DeepSpeed (Microsoft)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FDeepSpeed) | Deep learning optimization library with inference acceleration. Enables running larger models on limited hardware through ZeRO optimization. |\n| [AirLLM](https:\u002F\u002Fgithub.com\u002Flyogavin\u002Fairllm) | Run large models (70B+) on consumer hardware with limited memory. Loads models layer-by-layer for extreme memory efficiency. |\n| [AI Toolkit for VS Code (Microsoft)](https:\u002F\u002Fmarketplace.visualstudio.com\u002Fitems?itemName=ms-windows-ai-studio.windows-ai-studio) | VS Code extension to browse, test, fine-tune, and deploy models locally. Integrates ONNX and llama.cpp. |\n| [Ollama Grid Search](https:\u002F\u002Fgithub.com\u002Fdezoito\u002Follama-grid-search) | Desktop utility for systematic model evaluation. Test multiple models, prompts, and inference parameters side-by-side via a Rust\u002FReact GUI. |\n\n---\n\n## 💬 AI Chatbot UIs\n\n> 📅 Last checked: June 8, 2026\n\nFree, open-source web interfaces for chatting with AI models — self-host or use hosted versions.\n\n| Name | Description |\n|------|-------------|\n| [Open WebUI](https:\u002F\u002Fopenwebui.com\u002F) | Feature-rich ChatGPT-like interface for Ollama and OpenAI-compatible backends. RAG, image generation, multi-user. [GitHub](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui) |\n| [LibreChat](https:\u002F\u002Flibrechat.ai\u002F) | Open-source ChatGPT clone supporting 40+ providers, multi-user, plugins, and RAG. [GitHub](https:\u002F\u002Fgithub.com\u002Fdanny-avila\u002FLibreChat) |\n| [AnythingLLM](https:\u002F\u002Fanythingllm.com\u002F) | All-in-one desktop app for chatting with documents and models. Built-in RAG pipeline. [GitHub](https:\u002F\u002Fgithub.com\u002FMintplex-Labs\u002Fanything-llm) |\n| [Big-AGI](https:\u002F\u002Fbig-agi.com\u002F) | Feature-rich AI chat with personas, multi-model support, voice, and code execution. [GitHub](https:\u002F\u002Fgithub.com\u002Fenricoros\u002Fbig-agi) |\n| [Lobe Chat](https:\u002F\u002Flobehub.com\u002F) | Modern, extensible chat framework with plugin system and multi-provider support. [GitHub](https:\u002F\u002Fgithub.com\u002Flobehub\u002Flobe-chat) |\n| [Chatbot UI](https:\u002F\u002Fwww.chatbotui.com\u002F) | Simple, clean ChatGPT interface. Easy to self-host with any OpenAI-compatible API. [GitHub](https:\u002F\u002Fgithub.com\u002Fmckaywrigley\u002Fchatbot-ui) |\n| [NextChat (ChatGPT-Next-Web)](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb\u002FNextChat) | Lightweight cross-platform chat app. Self-host on Vercel or download official desktop\u002Fmobile clients. |\n\n---\n\n## 🤖 AI Coding Assistants\n\n> 📅 Last checked: June 8, 2026\n\nFree tools that integrate AI into your development workflow.\n\n| Name | Description |\n|------|-------------|\n| [Continue.dev](https:\u002F\u002Fcontinue.dev\u002F) | Open-source AI code assistant for VS Code and JetBrains. Chat, autocomplete, and edit with any model. [GitHub](https:\u002F\u002Fgithub.com\u002Fcontinuedev\u002Fcontinue) |\n| [Aider](https:\u002F\u002Faider.chat\u002F) | AI pair programming in the terminal. Edits code in your local git repo. Supports GPT, Claude, and local models. [GitHub](https:\u002F\u002Fgithub.com\u002Fpaul-gauthier\u002Faider) |\n| [Codeium (Windsurf)](https:\u002F\u002Fcodeium.com\u002F) | Free AI code assistant with autocomplete, chat, and search. Individual plan is free forever. |\n| [Tabby](https:\u002F\u002Ftabby.tabbyml.com\u002F) | Self-hosted AI coding assistant with no dependency on external services. [GitHub](https:\u002F\u002Fgithub.com\u002FTabbyML\u002Ftabby) |\n| [Cody (Sourcegraph)](https:\u002F\u002Fsourcegraph.com\u002Fcody) | Free tier for individuals. Chat, autocomplete, and commands with codebase context. |\n| [Llama Coder (Nutlope)](https:\u002F\u002Fllamacoder.together.ai\u002F) | Free AI code generation tool. Generate entire apps from prompts. |\n| [Bolt.new (StackBlitz)](https:\u002F\u002Fbolt.new\u002F) | Free tier for AI-powered full-stack web app development in browser. |\n| [Claude Code (Anthropic)](https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code\u002Foverview) | Free tier with limited usage for terminal-based AI coding assistant. |\n| [Cursor 3](https:\u002F\u002Fwww.cursor.com\u002F) | **Apr 2026.** AI-native code editor with deep model integration and agentic features. Free tier available. |\n| [OpenCode](https:\u002F\u002Fopencode.ai\u002F) | Go-based terminal AI coding assistant. Model-neutral, supports multiple LLM providers, LSP integration, and MCP tools. [GitHub](https:\u002F\u002Fgithub.com\u002Fopencode-ai\u002Fopencode) |\n| [CodeBuff](https:\u002F\u002Fwww.codebuff.com\u002F) | CLI-based AI coding assistant that understands entire codebases. Multi-agent architecture, works with any model provider through natural language instructions. |\n| [Pi](https:\u002F\u002Fpi.dev\u002F) | Open-source terminal AI coding agent with a unified multi-provider API. Model-agnostic, supports OpenAI, Anthropic, Google, and any OpenAI-compatible endpoint. Extensible plugin architecture. [GitHub](https:\u002F\u002Fgithub.com\u002Fearendil-works\u002Fpi) |\n| [Cline](https:\u002F\u002Fcline.bot\u002F) | Popular autonomous VS Code agent. Creates\u002Fedits files, runs terminal commands, browses web. Open-source, BYOK (bring your own API key). [GitHub](https:\u002F\u002Fgithub.com\u002Fcline\u002Fcline) |\n| [Roo Code](https:\u002F\u002Fgithub.com\u002FRooVetGit\u002FRoo-Code) | Community fork of Cline with faster feature releases. Open-source VS Code agent with deep model integration. |\n| [OpenHands](https:\u002F\u002Fall-hands.dev\u002F) | Autonomous AI software engineer. Navigates file systems, runs shell commands, tests code in browser. Self-hostable. [GitHub](https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands) |\n| [Twinny](https:\u002F\u002Fgithub.com\u002Ftwinnydotdev\u002Ftwinny) | Local-first AI coding extension for VS Code. Works entirely offline with local LLMs (Ollama, llama.cpp). Zero external dependencies. |\n| [Kodu (Claude Coder)](https:\u002F\u002Fgithub.com\u002Fkodu-ai\u002Fclaude-coder) | VS Code autonomous coding agent. Builds projects from scratch, handles complex tasks with natural language. |\n| [Goose](https:\u002F\u002Fblock.github.io\u002Fgoose\u002F) | Open-source CLI agent for complex software engineering tasks. Extensible plugin system. Built by Block\u002FSquare. [GitHub](https:\u002F\u002Fgithub.com\u002Fblock\u002Fgoose) |\n\n---\n\n## 📝 Code Models\n\n> 📅 Last checked: June 8, 2026\n\nSpecialized for code generation, completion, and analysis.\n\n| Name | Description |\n|------|-------------|\n| [MAI-Code-1-Flash (Microsoft)](https:\u002F\u002Fhuggingface.co\u002Fmicrosoft) | **Jun 2026.** Microsoft's open-weight coding model for lowering infrastructure costs. |\n| [DeepSeek Coder](https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai) | State-of-the-art open-weight code generation. DeepSeek's coder series leads SWE-bench. MIT license. |\n| [Qwen2.5-Coder (Alibaba)](https:\u002F\u002Fhuggingface.co\u002Fcollections\u002FQwen\u002Fqwen25-coder) | Highly capable code model series (1.5B–32B). Excellent balance of speed and quality. Apache 2.0. |\n| [Codestral (Mistral)](https:\u002F\u002Fhuggingface.co\u002Fmistralai\u002FCodestral-22B-v0.1) | Mistral's dedicated code generation model — fill-in-the-middle, completion, and instruction. [GitHub](https:\u002F\u002Fgithub.com\u002Fmistralai\u002Fcodestral) |\n| [CodeGemma (Google)](https:\u002F\u002Fhuggingface.co\u002Fgoogle\u002Fcodegemma-7b) | Google's Gemma architecture fine-tuned for code completion and instruction. Apache 2.0. |\n| [StarCoder2 (BigCode)](https:\u002F\u002Fhuggingface.co\u002Fbigcode\u002Fstarcoder2-15b) | Transparently trained code model covering 619 languages. OpenRAIL-M license. |\n| [Yi-Coder (01.AI)](https:\u002F\u002Fhuggingface.co\u002F01-ai\u002FYi-Coder-9B-Chat) | Efficient coding model with strong long-context understanding. Yi License (Apache 2.0 compatible). |\n| [Granite Code (IBM)](https:\u002F\u002Fhuggingface.co\u002Fibm-granite\u002Fgranite-8b-code) | IBM's enterprise-grade code model, available in multiple sizes. Apache 2.0. |\n| [Phi-4-mini (Microsoft)](https:\u002F\u002Fhuggingface.co\u002Fmicrosoft\u002FPhi-4-mini-instruct) | Lightweight model optimized for reasoning and code. Punches above its weight class. MIT license. |\n| [Qwen3-Coder-Next (Alibaba)](https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen3-Coder-Next) | **Early 2026.** Latest generation of Qwen's code series. Strong reasoning and long-context coding capabilities. Apache 2.0. |\n| [CodeLlama (Meta)](https:\u002F\u002Fhuggingface.co\u002Fmeta-llama\u002FCodeLlama-7b-hf) | **Aug 2023.** Llama 2-based code generation pioneer. Supports infilling, completion, and instruction. Llama 2 Community License. |\n| [WizardCoder (WizardLM)](https:\u002F\u002Fhuggingface.co\u002FWizardLMTeam\u002FWizardCoder-15B-V1.0) | **2023.** Evol-Instruct fine-tuned for complex coding tasks. Strong general code generation performance. Apache 2.0. |\n| [OpenCodeInterpreter](https:\u002F\u002Fhuggingface.co\u002Fm-a-p\u002FOpenCodeInterpreter-DS-6.7B) | **2024.** Integrates execution feedback to iteratively improve generated code. Bridges generation and execution. Apache 2.0. |\n| [Stable Code 3B (Stability AI)](https:\u002F\u002Fhuggingface.co\u002Fstabilityai\u002Fstable-code-3b) | **Aug 2023.** Lightweight 3B code model optimized for fill-in-the-middle. Efficient for local autocompletion. StabilityAI license. |\n| [CodeGeeX2 (THUDM)](https:\u002F\u002Fhuggingface.co\u002FTHUDM\u002Fcodegeex2-6b) | **2023.** Multilingual code model supporting 20+ languages. Strong in both Chinese and English code tasks. Apache 2.0. |\n| [CodeT5+ (Salesforce)](https:\u002F\u002Fhuggingface.co\u002FSalesforce\u002Fcodet5p-16b) | **2023.** Encoder-decoder architecture unifying code generation, completion, and understanding. BSD-3 license. |\n| [SantaCoder (BigCode)](https:\u002F\u002Fhuggingface.co\u002Fbigcode\u002Fsantacoder) | **2023.** Light 1.1B model specialized for Python, Java, and JavaScript. Fast and efficient for IDE integration. |\n\n---\n\n## 🔍 RAG & Vector Databases\n\n> 📅 Last checked: June 8, 2026\n\nFree tools for building retrieval-augmented generation pipelines — vector storage, embedding search, and document retrieval.\n\n| Name | Description |\n|------|-------------|\n| [Chroma](https:\u002F\u002Fwww.trychroma.com\u002F) | AI-native open-source embedding database. Runs in-process, no GPU needed. [GitHub](https:\u002F\u002Fgithub.com\u002Fchroma-core\u002Fchroma) |\n| [Qdrant](https:\u002F\u002Fqdrant.tech\u002F) | High-performance vector search engine. Free tier on Qdrant Cloud or self-host via Docker. [GitHub](https:\u002F\u002Fgithub.com\u002Fqdrant\u002Fqdrant) |\n| [pgvector](https:\u002F\u002Fgithub.com\u002Fpgvector\u002Fpgvector) | Vector similarity search inside PostgreSQL. Free if you already run Postgres. |\n| [LanceDB](https:\u002F\u002Flancedb.com\u002F) | Developer-friendly vector database built on Lance columnar format. Runs locally, no server needed. [GitHub](https:\u002F\u002Fgithub.com\u002Flancedb\u002Flancedb) |\n| [Weaviate](https:\u002F\u002Fweaviate.io\u002F) | Open-source vector database. Free sandbox tier on Weaviate Cloud. [GitHub](https:\u002F\u002Fgithub.com\u002Fweaviate\u002Fweaviate) |\n| [Milvus (Zilliz)](https:\u002F\u002Fmilvus.io\u002F) | Cloud-native vector database. Free tier on Zilliz Cloud or self-host. [GitHub](https:\u002F\u002Fgithub.com\u002Fmilvus-io\u002Fmilvus) |\n| [txtai](https:\u002F\u002Fneuml.github.io\u002Ftxtai\u002F) | AI-powered semantic search and RAG in a single Python package. [GitHub](https:\u002F\u002Fgithub.com\u002Fneuml\u002Ftxtai) |\n| [R2R (SciPhi)](https:\u002F\u002Fgithub.com\u002FSciPhi-AI\u002FR2R) | Production-ready RAG engine with API, user management, and observability. |\n| [Docling (IBM)](https:\u002F\u002Fwww.docling.ai\u002F) | Document understanding and conversion for RAG pipelines. Extracts PDFs, images, and more. [GitHub](https:\u002F\u002Fgithub.com\u002FDS4SD\u002Fdocling) |\n| [Unstructured.io](https:\u002F\u002Funstructured.io\u002F) | Preprocessing toolkit for documents (PDF, HTML, Word) for RAG pipelines. Free tier available. |\n\n---\n\n## 🧩 Agentic Frameworks\n\n> 📅 Last checked: June 8, 2026\n\nFree, open-source frameworks for building AI agents and multi-agent systems.\n\n| Name | Description |\n|------|-------------|\n| [LangGraph (LangChain)](https:\u002F\u002Flangchain-ai.github.io\u002Flanggraph\u002F) | Low-level framework for building stateful, multi-agent applications. [GitHub](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph) |\n| [CrewAI](https:\u002F\u002Fwww.crewai.com\u002F) | Multi-agent framework for orchestrating specialized AI agents to work together. [GitHub](https:\u002F\u002Fgithub.com\u002FcrewAIInc\u002FcrewAI) |\n| [AutoGen (Microsoft)](https:\u002F\u002Fmicrosoft.github.io\u002Fautogen\u002F) | Extensible framework for building multi-agent conversations. [GitHub](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fautogen) |\n| [Agno (formerly Phidata)](https:\u002F\u002Fwww.agno.com\u002F) | Full-stack AI framework for building multimodal agents with memory, knowledge, and tools. [GitHub](https:\u002F\u002Fgithub.com\u002Fagno-org\u002Fagno) |\n| [PydanticAI](https:\u002F\u002Fai.pydantic.dev\u002F) | Agent framework by Pydantic with type-safe outputs and dependency injection. [GitHub](https:\u002F\u002Fgithub.com\u002Fpydantic\u002Fpydantic-ai) |\n| [Mastra](https:\u002F\u002Fmastra.ai\u002F) | TypeScript framework for building AI applications and agent workflows. [GitHub](https:\u002F\u002Fgithub.com\u002Fmastra-ai\u002Fmastra) |\n| [OpenAI Agents SDK](https:\u002F\u002Fopenai.github.io\u002Fopenai-agents-python\u002F) | Lightweight SDK for building single and multi-agent systems. [GitHub](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-agents-python) |\n| [Semantic Kernel (Microsoft)](https:\u002F\u002Flearn.microsoft.com\u002Fen-us\u002Fsemantic-kernel\u002F) | SDK for orchestrating AI agents with planners, memory, and connectors. [GitHub](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fsemantic-kernel) |\n| [Dify](https:\u002F\u002Fdify.ai\u002F) | LLM app development platform with visual workflow builder and agent capabilities. [GitHub](https:\u002F\u002Fgithub.com\u002Flanggenius\u002Fdify) |\n| [Flowise](https:\u002F\u002Fflowiseai.com\u002F) | Low-code visual LLM flow builder with drag-and-drop interface. [GitHub](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) |\n| [TaskWeaver (Microsoft)](https:\u002F\u002Fmicrosoft.github.io\u002FTaskWeaver\u002F) | Code-first agent framework for planning and executing complex tasks. [GitHub](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FTaskWeaver) |\n| [Fazm](https:\u002F\u002Fgithub.com\u002Fmediar-ai\u002Ffazm) | **Apr 2026.** Open-source local computer-use agent for macOS. Drives apps via accessibility APIs, model-agnostic, faster than screenshot-based agents. |\n| [Smolagents (Hugging Face)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsmolagents) | Minimalist agent library where agents \"think in code.\" Lightweight, zero boilerplate. Supports code agents and tool-calling agents. |\n| [Swarms](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Fswarms) | Enterprise-grade multi-agent orchestration framework. Scalable infrastructure for autonomous agent swarms. Highly modular. |\n| [Letta (MemGPT)](https:\u002F\u002Fgithub.com\u002Fletta-ai\u002Fletta) | Framework for long-term agent memory. Virtual memory management that pages data in\u002Fout of context like an OS. Persistent agents. |\n| [Griptape](https:\u002F\u002Fgithub.com\u002Fgriptape-ai\u002Fgriptape) | Enterprise agent framework with strictly typed Pipelines, Workflows, and Agents. Structure-first, production-ready. |\n| [OpenAI Swarm](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm) | Experimental lightweight multi-agent orchestration. Uses Agents and Handoffs abstractions. Educational and minimalist. |\n| [Atomic Agents](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI\u002Fatomic-agents) | Framework inspired by Atomic Design. Compose agents from small, reusable, modular components. Testable and scalable. |\n| [PraisonAI](https:\u002F\u002Fgithub.com\u002FMervinPraison\u002FPraisonAI) | Low-code multi-agent framework. Define agent roles, tasks, and flows via YAML configuration. Wraps underlying agent frameworks. |\n| [Cognee](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee) | GraphRAG framework for agent knowledge management. Builds interconnected knowledge graphs from unstructured data. |\n| [AgentZero](https:\u002F\u002Fgithub.com\u002FAgentzerodotfun\u002Fagent-zero-main) | Self-healing autonomous agent with web UI. Manages own workflows, tool use, and environment. Self-evolving capabilities. |\n| [MetaGPT](https:\u002F\u002Fgithub.com\u002Fgeekan\u002FMetaGPT) | Multi-agent framework simulating a full software team. Assigns Agent, Product Manager, Engineer roles. Implements SOPs for end-to-end code generation. |\n| [ChatDev (OpenBMB)](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FChatDev) | Virtual software company driven by multi-agent collaboration. Follows waterfall model through design, coding, testing, and documentation. |\n| [AutoGPT](https:\u002F\u002Fgithub.com\u002FSignificant-Gravitas\u002FAutoGPT) | The original autonomous agent experiment. Sets its own goals, iterates on tasks, and executes without continuous human input. Web browsing and file management. |\n| [Bee Agent Framework (IBM)](https:\u002F\u002Fgithub.com\u002Fi-am-bee\u002Fbeeai-framework) | Production-ready framework for building reliable AI agents in Python and TypeScript. Modular, with built-in observability and IBM research optimizations. |\n| [Eliza (ai16z)](https:\u002F\u002Fgithub.com\u002Fai16z\u002Feliza) | Multi-platform agent framework for creating character-driven AI agents. Handles social media interaction, complex decision-making, and autonomous behavior across platforms. |\n| [SuperAGI](https:\u002F\u002Fgithub.com\u002FTransformerOptimus\u002FSuperAGI) | Developer-focused autonomous agent platform with GUI. Built-in resource management, file handling, and multi-tasking for running agents at scale. |\n| [AgentVerse (OpenBMB)](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FAgentVerse) | Framework for building and evaluating multi-agent environments. Easily configure agent teams and measure collaborative performance. |\n| [Qwen-Agent (Alibaba)](https:\u002F\u002Fgithub.com\u002FQwenLM\u002FQwen-Agent) | Agent framework tightly integrated with the Qwen model family. Optimized for function calling, code execution, RAG, and tool use with Qwen models. |\n| [AGiXT](https:\u002F\u002Fgithub.com\u002FJosh-XT\u002FAGiXT) | Extensible modular AI agent automation platform. Plugin system for swapping LLMs, memory backends, and tools. Highly customizable agent workflows. |\n\n---\n\n## 🎛 Fine-tuning Tools\n\n> 📅 Last checked: June 8, 2026\n\nTools to fine-tune free models on your own data — all free and open-source.\n\n| Name | Description |\n|------|-------------|\n| [Unsloth](https:\u002F\u002Fgithub.com\u002Funslothai\u002Funsloth) | Fast memory-efficient fine-tuning. 2x faster, 50% less memory. Supports QLoRA, LoRA, full fine-tune. |\n| [Axolotl](https:\u002F\u002Fgithub.com\u002FOpenAccess-AI-Collective\u002Faxolotl) | Streamlined fine-tuning framework supporting multiple model architectures and quantization methods. |\n| [LLaMA-Factory](https:\u002F\u002Fgithub.com\u002Fhiyouga\u002FLLaMA-Factory) | Easy-to-use fine-tuning with web UI. Supports 100+ models, multiple training methods. |\n| [Hugging Face TRL](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftrl) | Transformer Reinforcement Learning library. SFT, PPO, DPOTrainer, GRPOTrainer for aligning models. |\n| [TorchTune (Meta)](https:\u002F\u002Fgithub.com\u002Fpytorch\u002Ftorchtune) | Native PyTorch library for fine-tuning LLMs. Simple, extensible, efficient. |\n| [AutoTrain (Hugging Face)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fautotrain-advanced) | No-code fine-tuning platform. Train models with a web UI or API. |\n| [XTuner (InternLM)](https:\u002F\u002Fgithub.com\u002FInternLM\u002Fxtuner) | Efficient fine-tuning toolkit supporting QLoRA, LoRA, and full fine-tune with multiple model architectures. |\n| [Ludwig (Predibase)](https:\u002F\u002Fludwig.ai\u002F) | Declarative ML framework. Fine-tune models with a simple config file. [GitHub](https:\u002F\u002Fgithub.com\u002Fludwig-ai\u002Fludwig) |\n\n---\n\n## ✨ Prompt Engineering Tools\n\n> 📅 Last checked: June 8, 2026\n\nFree tools for testing, managing, and optimizing prompts.\n\n| Name | Description |\n|------|-------------|\n| [Promptfoo](https:\u002F\u002Fwww.promptfoo.dev\u002F) | Open-source tool for prompt testing and evaluation. Systematic A\u002FB testing of prompts. [GitHub](https:\u002F\u002Fgithub.com\u002Fpromptfoo\u002Fpromptfoo) |\n| [Fabric (Daniel Miessler)](https:\u002F\u002Fgithub.com\u002Fdanielmiessler\u002Ffabric) | Open-source framework for augmenting humans with AI. Library of curated prompts (patterns) for common tasks. |\n| [LangFuse](https:\u002F\u002Flangfuse.com\u002F) | Open-source LLM engineering platform with prompt management, versioning, and evaluation. [GitHub](https:\u002F\u002Fgithub.com\u002Flangfuse\u002Flangfuse) |\n| [OpenPrompt (THUNLP)](https:\u002F\u002Fthunlp.github.io\u002FOpenPrompt\u002F) | Framework for prompt-learning research. Supports template and verbalizer design. [GitHub](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FOpenPrompt) |\n| [DSPy (Stanford)](https:\u002F\u002Fdspy.ai\u002F) | Framework for algorithmically optimizing LM prompts and weights. [GitHub](https:\u002F\u002Fgithub.com\u002Fstanfordnlp\u002Fdspy) |\n| [Agenta](https:\u002F\u002Fagenta.ai\u002F) | Open-source LLM platform for prompt management, evaluation, and deployment. [GitHub](https:\u002F\u002Fgithub.com\u002FAgenta-AI\u002Fagenta) |\n\n---\n\n## 📊 Datasets\n\n> 📅 Last checked: June 8, 2026\n\nFree, open datasets for training, fine-tuning, and evaluating models.\n\n| Name | Description |\n|------|-------------|\n| [Hugging Face Datasets](https:\u002F\u002Fhuggingface.co\u002Fdatasets) | The standard hub for open datasets. 150,000+ datasets across all tasks. |\n| [Common Corpus](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fcommon-catalog) | Massive open-source dataset for training large language models. |\n| [The Stack v2 (BigCode)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fbigcode\u002Fthe-stack-v2) | Large-scale code dataset covering 619 programming languages. Permissive license. |\n| [FineWeb (Hugging Face)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FHuggingFaceFW\u002Ffineweb) | High-quality web dataset for LLM pre-training. 15T tokens. |\n| [Dolly (Databricks)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fdatabricks\u002Fdatabricks-dolly-15k) | 15k instruction-response pairs for fine-tuning. CC-BY-SA. |\n| [OpenAssistant Conversations](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FOpenAssistant\u002Foasst1) | 160k human-generated assistant conversations. Apache 2.0. |\n| [ShareGPT (RyokoAI)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fanon8231489123\u002FShareGPT_Vicuna_unfiltered) | Real user-ChatGPT conversations for fine-tuning. |\n| [UltraChat (Sean C.)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FHuggingFaceH4\u002Fultrachat_200k) | 200k multi-turn conversations synthesized by ChatGPT. |\n| [No Robots (Hugging Face)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FHuggingFaceH4\u002Fno_robots) | 10k high-quality human-written instructions. Apache 2.0. |\n| [RLAIF-V (OpenBMB)](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fopenbmb\u002FRLAIF-V) | AI-generated preference data for RLHF. Apache 2.0. |\n| [MMLU \u002F GSM8K](https:\u002F\u002Fhuggingface.co\u002Fdatasets) | Standard benchmarks for evaluation. |\n\n---\n\n## ☁ Model Hosting Platforms\n\n> 📅 Last checked: June 8, 2026\n\nFree platforms that host models — run inference without downloading anything.\n\n| Name | Description |\n|------|-------------|\n| [Hugging Face Spaces](https:\u002F\u002Fhuggingface.co\u002Fspaces) | Free hosting for ML apps (Gradio, Streamlit). Thousands of community demos. |\n| [Hugging Face Inference Endpoints (Free Tier)](https:\u002F\u002Fhuggingface.co\u002Finference-endpoints) | Deploy models with free trial credits. |\n| [Google Colab (Free Tier)](https:\u002F\u002Fcolab.research.google.com\u002F) | Free GPU (T4, sometimes A100). Perfect for running models and fine-tuning. |\n| [Kaggle Notebooks](https:\u002F\u002Fwww.kaggle.com\u002Fcode) | Free GPU (T4 x2). 30 hours\u002Fweek. Good for heavier workloads. |\n| [Lightning AI Studio](https:\u002F\u002Flightning.ai\u002F) | Free tier with GPU access for development and prototyping. |\n| [Modal](https:\u002F\u002Fmodal.com\u002F) | Free monthly credits for serverless GPU compute. |\n| [Replicate (Free Tier)](https:\u002F\u002Freplicate.com\u002F) | Free credits for running community models. |\n| [Deepnote](https:\u002F\u002Fdeepnote.com\u002F) | Free tier with GPU for data science and ML notebooks. |\n\n---\n\n## 📚 Learning Resources\n\n> 📅 Last checked: June 8, 2026\n\nFree courses, books, and tutorials for learning AI and LLMs.\n\n| Name | Description |\n|------|-------------|\n| [Fast.ai](https:\u002F\u002Fwww.fast.ai\u002F) | Code-first deep learning education. Practical, free courses from fundamentals to advanced. |\n| [Hugging Face NLP Course](https:\u002F\u002Fhuggingface.co\u002Flearn\u002Fnlp-course) | Comprehensive free course on transformers, tokenizers, datasets, and deployment. |\n| [DeepLearning.AI Short Courses](https:\u002F\u002Fwww.deeplearning.ai\u002Fshort-courses\u002F) | Free short courses on LLMs, RAG, LangChain, and AI agents. |\n| [Full Stack Deep Learning](https:\u002F\u002Ffullstackdeeplearning.com\u002F) | Free course on ML engineering: training, deploying, and maintaining models. |\n| [Andrej Karpathy's Course](https:\u002F\u002Fkarpathy.ai\u002Fzero-to-hero.html) | From-scratch neural network implementation videos. |\n| [Neural Networks: Zero to Hero](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=VMj-3S1tku0&list=PLAqhIrjkxbuWI23v9cThsA9GvCAUhRvKZ) | YouTube series building neural networks from scratch. |\n| [LLM University (Cohere)](https:\u002F\u002Fdocs.cohere.com\u002Fdocs\u002Fllmu) | Free course on LLMs, embeddings, and RAG. |\n| [Prompt Engineering Guide (DAIR.AI)](https:\u002F\u002Fwww.promptingguide.ai\u002F) | Comprehensive free guide on prompt engineering techniques. |\n| [Anthropic Cookbook](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fanthropic-cookbook) | Free recipes and patterns for working with Claude. |\n| [OpenAI Cookbook](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-cookbook) | Free examples and guides for the OpenAI API. |\n\n---\n\n## 🏆 Resources & Leaderboards\n\n> 📅 Last checked: June 8, 2026\n\n| Name | Description |\n|------|-------------|\n| [Perplexity](https:\u002F\u002Fwww.perplexity.ai\u002F) | Free AI search and research assistant with real-time answers and source citations. |\n| [Hugging Face Open LLM Leaderboard](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002Fopen-llm-leaderboard\u002Fopen_llm_leaderboard) | The primary benchmark for open-weight models. Updated regularly. |\n| [LMSYS Chatbot Arena](https:\u002F\u002Flmarena.ai\u002F) | Human preference rankings of models. Best source for real-world quality comparisons. |\n| [Artificial Analysis](https:\u002F\u002Fartificialanalysis.ai\u002F) | Independent benchmarks for speed, pricing, and quality across providers. |\n| [Hugging Face Models](https:\u002F\u002Fhuggingface.co\u002Fmodels) | Search 1M+ models. Filter by license, task, framework. |\n| [OpenRouter Models](https:\u002F\u002Fopenrouter.ai\u002Fmodels) | Browse models available via API with pricing and free tiers. |\n| [Ollama Library](https:\u002F\u002Follama.com\u002Flibrary) | Browse models available for one-command local setup. |\n| [cheahjs\u002Ffree-llm-api-resources](https:\u002F\u002Fgithub.com\u002Fcheahjs\u002Ffree-llm-api-resources) | Community-maintained list of free LLM API resources. |\n| [SweetTea](https:\u002F\u002Fwww.sweettea.ai\u002F) | Community voting on model quality and preference. |\n\n---\n\n## 👥 Communities\n\n> 📅 Last checked: June 8, 2026\n\n| Name | Description |\n|------|-------------|\n| [Hugging Face Discord](https:\u002F\u002Fdiscord.gg\u002Fhuggingface) | Model releases, discussions, and community support. |\n| [r\u002FLocalLLaMA](https:\u002F\u002Freddit.com\u002Fr\u002FLocalLLaMA) | The largest Reddit community for running local LLMs. |\n| [Ollama Discord](https:\u002F\u002Fdiscord.gg\u002Follama) | Ollama community for local model enthusiasts. |\n| [LM Studio Discord](https:\u002F\u002Fdiscord.gg\u002Flmstudio) | LM Studio community. |\n| [Hugging Face Forums](https:\u002F\u002Fdiscuss.huggingface.co\u002F) | Discussions on models, datasets, and Spaces. |\n| [r\u002FMachineLearning](https:\u002F\u002Freddit.com\u002Fr\u002FMachineLearning) | General ML\u002FAI research and news. |\n| [Discord: AI Agents](https:\u002F\u002Fdiscord.gg\u002Fai-agents) | Community for AI agent development and agentic frameworks. |\n\n---\n\n## License\n\n[![CC0](https:\u002F\u002Fmirrors.creativecommons.org\u002Fpresskit\u002Fbuttons\u002F88x31\u002Fsvg\u002Fcc-zero.svg)](https:\u002F\u002Fcreativecommons.org\u002Fpublicdomain\u002Fzero\u002F1.0\u002F)\n\nTo the extent possible under law, the author has waived all copyright and related or neighboring rights to this work.\n","2026-06-11 04:11:55","CREATED_QUERY"]