Awesome AI 
A curated list of AI APIs, tools, platforms, and learning resources.
⚠️ = archived or without repository activity for more than 12 months; verify status before adoption.
How to Choose
🧠 LLM API Providers
| Tool |
Description |
| OpenAI API |
GPT-4o and o-series reasoning models with tool use, vision, and multimodal capabilities. |
| Anthropic Claude API |
Claude model family with 200K context windows, vision, and agentic capabilities; safety-focused. |
| Google Gemini API |
Gemini models with 1M+ token context, search integration, and Google Cloud ecosystem. |
| Mistral AI API |
European AI lab with open-weight and commercial models under Apache 2.0 licensing at competitive prices. |
| DeepSeek API |
DeepSeek V3 and R1 reasoning models at a fraction of competitor pricing with 128K context. |
| Cohere API |
Command R+ models optimized for RAG, multilingual support, and enterprise deployment. |
| AI21 Labs API |
Jamba models and Maestro orchestration system for enterprise NLP with strong multilingual capabilities. |
| xAI Grok API |
Grok models with competitive pricing, image generation, and video generation. |
| Inflection AI |
Creators of Pi, focused on empathetic and helpful interactions via enterprise API. |
| Reka AI |
Multimodal models spanning text, images, and video understanding. |
(⬆ back to top)
🔓 Open Source LLM Models
| Tool |
Stars |
Description |
| Llama 4 (Meta) |
- |
Scout (17B active/109B experts) and Maverick variants, 128K context, strong general performance. |
| Qwen 3 (Alibaba) |
- |
MoE family (up to 235B parameters) excelling in multilingual and reasoning tasks with open weights. |
| DeepSeek V3 |
105k |
671B MoE model (MIT license) matching frontier models at a fraction of the cost. |
| DeepSeek R1 |
92k |
Reasoning model trained via large-scale RL achieving o1-level performance (MIT license). |
| Mistral Large / Mixtral |
- |
Mixtral MoE and Small 3 (24B) punching above their weight in efficiency. |
| Gemma 3 (Google) |
- |
Available in multiple sizes with strong performance for its parameter count. |
| Phi-4 (Microsoft) |
- |
Small language model built on synthetic datasets achieving state-of-the-art performance for its size. |
| Command R+ (Cohere) |
- |
Optimized for RAG and enterprise use with strong multilingual capabilities. |
| Yi (01.AI) |
- |
Bilingual (Chinese-English) models with strong performance in independent benchmarks. |
| StarCoder2 |
2k |
Code LLM in 3B/7B/15B sizes where the 15B matches 33B+ models on code evaluations. |
| OLMo (Allen AI) |
- |
Training data, code, and weights all publicly available for reproducible research. |
| DBRX (Databricks) |
- |
MoE model with 132B total / 36B active parameters for enterprise use cases. |
(⬆ back to top)
🤔 AI Reasoning Models
(⬆ back to top)
🚀 Open-Source Model Hosting & Inference
| Tool |
Description |
| Together AI |
200+ models with sub-100ms latency, fine-tuning, and horizontal scaling. |
| Fireworks AI |
Low-latency inference for open-weight models including DeepSeek, Llama, and Qwen. |
| Groq |
Purpose-built LPU hardware delivering ultra-fast inference with extremely low latency. |
| Cerebras |
Wafer-scale chip architecture eliminating GPU communication bottlenecks for fast inference. |
| SambaNova |
Enterprise inference with OpenAI-compatible API and ultra-low latency. |
| Replicate |
Run and deploy models via API with simple scaling across image, video, audio, and text. |
| Modal |
Serverless cloud for AI models with pay-per-use pricing and infrastructure control. |
| Baseten |
Deploy and serve ML models with infrastructure control and optimized performance. |
| DeepInfra |
Cost-effective inference API with managed servers, GPUs, and auto-scaling. |
| Lepton AI |
Pythonic framework for building and deploying AI services with cloud-native inference. |
| fal.ai |
600+ models with competitive pricing, optimized for generative media. |
| Novita AI |
Affordable access to models for text, image, and audio generation. |
(⬆ back to top)
🔀 LLM API Routers & Gateways
| Tool |
Stars |
Description |
| OpenRouter |
- |
Unified API for 100+ LLMs from multiple providers with transparent pricing comparison. |
| LiteLLM |
60k |
Proxy/gateway providing a unified API to 100+ LLMs with load balancing, fallbacks, and spend tracking. |
| Martian |
- |
AI-powered router that automatically selects the best model for each request using Model Mapping. |
| Portkey |
- |
Enterprise gateway with load balancing, caching, fallbacks, and observability. |
| Helicone |
- |
Rust-based gateway with load balancing, native observability, and cost tracking. |
(⬆ back to top)
☁️ Cloud AI Platforms (Managed)
| Tool |
Description |
| AWS Bedrock |
Multi-vendor model marketplace with AgentCore, supporting Anthropic, Meta, Cohere, and Amazon models. |
| Azure AI Foundry |
11,000+ models with GPT-4/5 access and deep Azure integration. |
| Google Vertex AI |
Gemini family, Model Garden for 3rd-party/OSS models, and data-driven MLOps. |
| AWS SageMaker |
End-to-end ML platform for building, training, and deploying models with managed infrastructure. |
| Azure Machine Learning |
Automated ML, responsible AI dashboard, and managed endpoints. |
| Databricks |
Unified analytics and AI with managed MLflow, Unity Catalog, and Mosaic AI. |
| IBM watsonx |
Foundation models, model training, and AI governance for business workflows. |
(⬆ back to top)
🤖 AI Agent Frameworks
| Tool |
Stars |
Description |
| LangChain |
- |
LLM application framework with extensive integrations for chains, agents, and retrieval. |
| LangGraph |
- |
Cyclical graphs for agent runtimes with state management and human-in-the-loop. Extends LangChain. |
| LlamaIndex |
- |
RAG-first framework with 100+ data connectors and purpose-built retrieval infrastructure. |
| CrewAI |
- |
Role-based multi-agent systems with structured task delegation. |
| Microsoft AutoGen |
- |
Multi-agent conversation framework merging with Semantic Kernel into unified Microsoft Agent Framework. |
| Microsoft Semantic Kernel |
- |
LLM integration SDK with deep Azure ecosystem support across C#, Python, and Java. |
| DSPy |
- |
Declarative framework for programming (not prompting) language models with automatic prompt optimization. |
| Haystack |
- |
AI orchestration framework for LLM applications with modular pipelines. |
| Instructor |
14k |
Structured output extraction using Pydantic models with validation and retry logic. |
| Pydantic AI |
- |
Agent framework from the Pydantic team with type-safe structured outputs. |
| OpenAI Agents SDK |
30k |
OpenAI's official framework for agentic workflows with tool use, handoffs, and guardrails. |
| Google ADK (Agent Development Kit) |
- |
Google's framework with native Gemini integration and multi-agent orchestration. |
| Smolagents |
30k |
Hugging Face's lightweight agent framework with minimal boilerplate. |
| CopilotKit |
- |
Framework for building AI copilots into React applications with AG-UI protocol support. |
(⬆ back to top)
💾 AI Memory & Context Management
Each tool takes a different approach to giving agents persistent memory across sessions.
| Tool |
Stars |
Description |
| Mem0 |
- |
Dedicated memory layer extracting and managing memories across sessions with up to 80% token reduction. |
| Zep |
- |
Temporal knowledge graph tracking how facts change over time with structured business data. |
| Letta (MemGPT) |
- |
Agent framework with first-class memory management exposing editable memory blocks and stateful runtime. |
| Cognee |
- |
Memory engine finding hidden connections in data using knowledge graphs with semantic embeddings. |
| LangMem |
2k |
Memory tools for LangGraph agents extracting information from conversations for long-term context. |
| Supermemory |
- |
Universal memory API with time-annotated semantic traces and scalable semantic recall. |
| Graphlit |
- |
Semantic content infrastructure with multimodal ingestion linked to a shared knowledge graph. |
(⬆ back to top)
📊 Structured Output & Data Extraction
| Tool |
Stars |
Description |
| Instructor |
14k |
Structured output extraction using Pydantic models with validation and retry logic. |
| BAML |
9k |
Domain-specific language for writing and testing LLM functions with type-safe validation. |
| Marvin |
6k |
Lightweight Python library for natural language interfaces that extract structured data from LLMs. |
| Outlines |
16k |
Constrained text generation using regular expressions, JSON schemas, and context-free grammars. |
| Guidance |
22k |
Microsoft's language for controlling LLMs with interleaving generation, prompting, and logical control. |
| SGLang |
37k |
Fast serving framework with built-in structured output via JSON schema, regex, and EBNF. |
| TypeChat |
9k |
Uses TypeScript types to guide LLM outputs into well-typed structured responses. |
| JSONFormer |
5k |
Constrains generation to only produce valid JSON tokens. |
| LlamaParse |
- |
Enterprise document parsing with agentic OCR for extracting structured data from complex documents. |
| Docling |
68k |
Document conversion toolkit supporting PDFs, office documents, HTML, images, and structured output. |
(⬆ back to top)
💻 AI Code Assistants
| Tool |
Description |
| GitHub Copilot |
20M+ users, deep IDE integration, agent mode, and enterprise features. |
| Cursor |
AI-native IDE built on VS Code with deep codebase understanding, multi-file editing, and Composer agent. |
| Windsurf (Codeium) |
Agentic IDE with Cascade flow combining deep context awareness with autonomous code generation and free tier. |
| Claude Code |
Terminal-based agentic coding excelling at reasoning, multi-file refactoring, and autonomous workflows. |
| Tabnine |
Privacy-first; multi-model support trained only on permissive licenses. |
| Amazon Q Developer |
Autonomous agents, AWS resource management, and console diagnostics. |
| Sourcegraph Cody |
Search-first architecture for analyzing entire multi-repo environments. |
| Continue.dev |
Supports any LLM and deep IDE customization. |
| Aider |
Free terminal-based AI pair programming tool for command-line workflows and automation. |
| Qodo (formerly CodiumAI) |
Focused on generating meaningful test suites and code review. |
| JetBrains AI |
Natively integrated into JetBrains IDEs with context-aware completions. |
| Google Gemini Code Assist |
Powered by Gemini models with Google Cloud integration. |
(⬆ back to top)
🎨 AI Image Generation APIs
| Tool |
Description |
| OpenAI DALL-E API |
Strong prompt understanding and seamless GPT integration. |
| Stability AI API |
Stable Diffusion 3.5 and SDXL models with fine-grained control and open-source weights. |
| Black Forest Labs (FLUX) |
FLUX image-generation and editing models available through an API. |
| Ideogram API |
Specializes in accurate text rendering, ideal for logos, posters, and branded content. |
| Recraft API |
Excels at vector outputs (icons, logos) alongside raster images with style controls. |
| Leonardo AI API |
Presets optimized for game art, textures, and product visuals. |
| Google Imagen API |
Imagen 3/4 models via Vertex AI with studio-quality outputs. |
| Midjourney |
Best for artistic and aesthetic outputs; accessible via Discord bot. |
(⬆ back to top)
🎬 AI Video Generation APIs
| Tool |
Description |
| Runway API |
Gen-4.5 with cinematic quality, motion brushes, and scene consistency. |
| Pika API |
PikaSwaps, Pikaffects, and fast rendering optimized for social media. |
| Luma AI (Dream Machine) |
4K HDR video with Ray3 engine and superior physics simulation. |
| Kling AI |
Up to 2-minute videos at 1080p with simultaneous audio-visual generation. |
| OpenAI Sora |
Cinematic-quality videos with realistic physics. |
| Synthesia API |
230+ AI avatars, 140+ languages, and Express-2 engine for enterprise video. |
| HeyGen API |
Realistic language dubbing with voice cloning and lip-sync. |
| D-ID API |
Talking-head videos from images and audio, supporting 120+ languages. |
| Google Veo API |
Native audio integration and high-fidelity output via Vertex AI. |
| Twelve Labs |
Multimodal video understanding: search, classification, and generation from video content. |
(⬆ back to top)
🎙️ AI Audio & Speech APIs
| Tool |
Description |
| ElevenLabs API |
3,000+ voices, 70+ languages, emotional nuance, and voice cloning. |
| OpenAI Audio API |
TTS and STT including gpt-4o-transcribe and Whisper for 50+ languages. |
| Deepgram API |
STT and TTS (Aura-2) optimized for conversational AI with sub-250ms latency. |
| AssemblyAI API |
STT with speaker diarization, sentiment analysis, and content moderation. |
| Cartesia API |
Ultra-low-latency voice generation (<150ms) with rapid voice cloning for interactive use. |
| Play.ht API |
900+ voices across 140+ languages and accents. |
| Resemble AI API |
Expressive speech synthesis, voice cloning, and deepfake detection for enterprise. |
| Speechify API |
Apple Design Award winner with 1,000+ voices in 60+ languages. |
| OpenAI Realtime API |
Speech-to-speech for real-time voice applications with low latency. |
(⬆ back to top)
🔍 AI Search APIs
| Tool |
Description |
| Perplexity Sonar API |
LLM-powered search providing direct answers with cited sources. |
| Tavily API |
Built for agents and RAG workflows with SOC 2 certification. |
| Exa API |
Semantic/neural search scoring 94.9% on SimpleQA, returning meaning-based results. |
| Brave Search API |
Independent index with no tracking and privacy-focused results. |
| SerpAPI |
SERP scraping across search engines for AI features, SEO, and price monitoring. |
| You.com API |
Deep Search, Express Agent, and MCP support. |
| Serper API |
Fast, affordable Google SERP API for high-volume use cases. |
| WebSearch API |
Web search API designed for AI workflow integration. |
(⬆ back to top)
🌐 AI Translation APIs
| Tool |
Description |
| DeepL API |
Known for natural, polished translations especially for European languages. |
| Google Cloud Translation |
130+ languages with AutoML custom model training. |
| Microsoft Translator |
100+ languages with custom translation models and document translation on Azure. |
| Amazon Translate |
Real-time and batch translation integrated into the AWS ecosystem. |
| ModernMT |
Adaptive engine that learns from human corrections in real-time. |
| Unbabel |
Blends AI translation with human editors for enterprise content. |
| LILT |
Adaptive neural models with human-in-the-loop workflow for enterprise localization. |
| Phrase |
Developer-friendly localization with extensive APIs and CLI tools for custom pipelines. |
| Lokalise |
400+ languages with tone, style, and industry customization. |
| Tolgee |
In-context translating features, popular with JavaScript developers. |
(⬆ back to top)
✍️ AI Writing & Content APIs
| Tool |
Description |
| Jasper API |
Content platform for marketing teams with brand voice, campaign management, and generation. |
| Writer API |
Full-stack content generation with brand governance and custom model training. |
| Copy.ai |
Marketing copy, sales emails, and Go-To-Market workflows. |
| Writesonic |
Fact-checked, SEO-friendly article generation pulling real-time data from Google. |
| Surfer SEO |
SERP analysis and real-time content scoring for SEO optimization. |
| Frase |
Analyzes SERPs in real-time with AI-driven content suggestions for SEO. |
| Grammarly API |
Grammar, tone, clarity, and brand-voice checking for integration into applications. |
| Anyword |
Predictive performance scoring that forecasts content performance before publishing. |
| Wordtune |
Rewriting and paraphrasing by AI21 Labs for improving clarity, tone, and conciseness. |
(⬆ back to top)
📐 Vector Databases & Embeddings
Vector Databases
| Tool |
Stars |
Description |
| Pinecone |
- |
Fully managed with serverless architecture and support for billions of vectors. |
| Weaviate |
- |
Excels at hybrid search (vector + keyword) with GraphQL API. |
| Qdrant |
- |
High QPS at 99% recall, production-proven at scale. |
| Chroma |
- |
Developer-friendly embedding database ideal for prototyping RAG systems. |
| Milvus |
- |
Leads in low-latency benchmarks with support for billions of vectors. |
| Zilliz Cloud |
- |
Managed cloud service for Milvus with enterprise features and production SLAs. |
| pgvector |
23k |
PostgreSQL extension keeping vectors and relational data together. |
| LanceDB |
- |
Serverless, built on Lance columnar format with zero-copy access. |
Embedding APIs
| Tool |
Description |
| OpenAI Embeddings API |
ada-002 and newer models for semantic search, clustering, and classification. |
| Voyage AI API |
Voyage 4 series with shared embedding spaces and domain-specific models. |
| Cohere Embed API |
Multilingual embeddings optimized for search and retrieval with Embed v3. |
| Google Embedding API |
Text and multimodal embeddings via Vertex AI with Gecko and newer models. |
| Jina AI Embeddings |
API-based embedding models with multilingual support and reranking. |
(⬆ back to top)
📚 RAG Tools & Frameworks
| Tool |
Description |
| LlamaIndex |
100+ data connectors, retrieval strategies, and production-ready pipelines. |
| Unstructured |
Converts PDFs, images, HTML, and Word docs into structured data for ingestion. |
| Vectara |
RAG-as-a-service with built-in semantic search, summarization, and hallucination detection. |
| Ragas |
Reference-free evaluation of RAG pipelines with metrics for context relevance and groundedness. |
| LangChain Retrievers |
Modular retrieval components supporting vector stores, multi-query, and contextual compression. |
| Cohere Rerank |
Neural reranking that improves search relevance by reordering results on semantic similarity. |
| Morphik |
RAG infrastructure for building and deploying retrieval-augmented generation systems. |
(⬆ back to top)
📄 AI Document Processing & OCR
| Tool |
Stars |
Description |
| Azure Document Intelligence |
- |
Extracts structure, relationships, and key-values from documents. |
| Google Document AI |
- |
Pre-trained models for invoices, receipts, and custom documents. |
| Amazon Textract |
- |
Extracts text, forms, and tables integrated with S3 and Lambda. |
| Mistral OCR |
- |
OCR API for extracting structured content from documents. |
| LlamaParse |
- |
Agentic OCR extracting structured data from complex documents. |
| Docling |
68k |
Document conversion toolkit with table, layout, and structured-output support. |
| Mathpix |
- |
Specialized in mathematical equations, scientific documents, and LaTeX conversion. |
| Reducto |
- |
API-first parsing for extracting structured data from PDFs for RAG pipelines. |
| Mindee |
- |
Developer-friendly extraction for invoices, receipts, IDs, and custom document types. |
(⬆ back to top)
🗃️ AI Database & SQL Assistants
| Tool |
Description |
| Vanna.ai |
Self-hosted AI SQL agent with personalized schema training. |
| Defog.ai |
Text-to-SQL engine achieving 90-95% accuracy, matching GPT-4 on specialized benchmarks. |
| Text2SQL.ai |
Converts natural language to SQL with multiple dialect support. |
| DBHub |
Universal database MCP server enabling AI assistants to query databases. |
| Outerbase |
Spreadsheet-like views with EZQL natural language agent. |
| SQL Chat |
Chat-based SQL client using natural language. |
| BlazeSQL |
Generates SQL from natural language by parsing schema metadata. |
| DataGPT |
Slack-native analytics answering business questions with raw SQL transparency. |
(⬆ back to top)
📏 LLM Evaluation & Monitoring
| Tool |
Description |
| LangSmith |
Tracing, prompt testing, and RAG assessment from the LangChain team. |
| Langfuse |
LLM engineering platform for tracing, prompt management, evaluations, and observability. |
| Braintrust |
Evaluation-first platform for systematic prompt experimentation with side-by-side comparison. |
| Arize AI |
LLM observability with Phoenix tracing and production monitoring. |
| Promptfoo |
Prompt engineering, A/B testing, and LLM-as-a-judge evaluations. |
| Humanloop |
Prompt management and evaluation with human feedback loops. |
| WhyLabs |
Detects anomalies, drift, hallucinations, and bias in production systems. |
| Galileo AI |
Luna-2 evaluation models and real-time guardrail system for agent actions. |
| TruLens |
Evaluating and tracing AI agents and RAG systems with OpenTelemetry tracing. |
| Maxim AI |
Testing, monitoring, and optimizing LLM systems. |
| Evidently AI |
Detects data drift and model performance degradation. |
| Fiddler AI |
Focused on explainability, fairness, and compliance for ML and LLM models. |
(⬆ back to top)
💰 AI Gateway & Cost Management
| Tool |
Description |
| Helicone |
Gateway and observability with request logging, cost tracking, and Rust-based performance. |
| Portkey |
Unified API with load balancing, semantic caching, and multi-provider fallbacks. |
| Keywords AI |
Tracks usage, latency, and spend across providers. |
| Lunary |
LLM monitoring and analytics with cost tracking and prompt management. |
| OpenRouter |
Unified API marketplace for 100+ LLMs with transparent pricing and cost optimization. |
(⬆ back to top)
🎯 Prompt Engineering & Management
| Tool |
Description |
| Langfuse |
Prompt management with versioning, tracing, evaluations, and observability. |
| PromptLayer |
Logging, versioning, visual editing, and regression testing. |
| Agenta |
LLMOps platform for prompt management, evaluation, and A/B testing. |
| Pezzo |
Lightweight prompt management for versioning, testing, and instant deployment. |
| Promptfoo |
CLI for prompt testing, evaluation, and red-teaming with YAML-based configuration. |
| Maxim AI |
Prompt management and evaluation with testing workflows and version control. |
(⬆ back to top)
🔧 Fine-Tuning Platforms
| Tool |
Stars |
Description |
| OpenAI Fine-Tuning |
- |
Managed fine-tuning for GPT models with reinforcement fine-tuning preview. |
| Together AI Fine-Tuning |
- |
Cloud-based fine-tuning for open-source models with LoRA adapter support. |
| Predibase |
- |
Low-code fine-tuning and hosting with LoRAX multi-adapter serving and serverless deployment. |
| LoRAX |
4k |
Serves 100s of fine-tuned LoRA adapters on a single GPU. |
| Modal Fine-Tuning |
- |
Serverless GPU compute with pay-per-use pricing and zero idle costs. |
| Lambda Cloud |
- |
GPU cloud optimized for AI training with zero egress fees and InfiniBand networking. |
| Anyscale |
- |
Built on Ray for distributed fine-tuning and serving of LLMs. |
(⬆ back to top)
🛡️ AI Safety & Guardrails
| Tool |
Stars |
Description |
| Guardrails AI |
- |
LLM output validation with pre-built validators from Guardrails Hub. |
| NVIDIA NeMo Guardrails |
7k |
Programmable safety guardrails using a domain-specific language. |
| Lakera Guard |
- |
Real-time prompt injection detection with customizable safeguards and monitoring dashboards. |
| LLM Guard |
- |
Scanners for prompt injection, PII, and toxicity in LLM interactions. |
| Rebuff |
2k |
Self-hardening prompt injection detector that learns from attack attempts over time. |
| Azure AI Content Safety |
- |
Detects harmful content in text and images with configurable severity levels. |
(⬆ back to top)
⚖️ AI Compliance & Governance
| Tool |
Description |
| Credo AI |
Compliance automation with EU AI Act, NIST AI RMF, and ISO 42001. |
| Holistic AI |
AI lifecycle governance assessing models for safety, fairness, bias, and regulatory alignment. |
| IBM watsonx.governance |
Model lifecycle management, bias detection, and compliance tracking. |
| Zenity |
AI agent compliance: risk detection, guardrail enforcement, and regulatory reporting. |
| Fairly AI |
AI risk management for financial services with model validation and bias detection. |
| Drata |
Compliance automation with AI-powered controls for SOC 2, ISO 27001, and AI regulations. |
| Securiti AI |
Data command center with AI governance, data privacy, and consent management. |
(⬆ back to top)
🏭 MLOps & Deployment
| Tool |
Description |
| MLflow |
Experiment tracking, model registry, and deployment; MLflow 3 adds GenAI ops support. |
| BentoML |
Build, ship, and scale AI model serving with high-performance inference. |
| KServe |
Kubernetes-native serverless model serving with autoscaling. |
| ZenML |
MLOps framework for portable ML pipelines across any infrastructure. |
| Metaflow |
Netflix-originated framework for real-life ML workflows with deep AWS integration. |
| NVIDIA Triton Inference Server |
GPU inference serving supporting multiple ML frameworks and model formats. |
| DVC (Data Version Control) |
Version control for data, models, and experiments alongside Git. |
| Kubeflow |
Google-backed ML toolkit for Kubernetes with pipeline orchestration and LLM fine-tuning. |
| Seldon Core |
ML model serving at scale with A/B testing and canary rollouts. |
| Weights & Biases |
Experiment tracking, model management, and dataset versioning for the full ML lifecycle. |
| Neptune.ai |
Experiment tracking and model registry for logging, comparing, and collaborating on models. |
| Comet ML |
Experiment management with real-time tracking, model comparison, and production monitoring. |
(⬆ back to top)
⚡ AI Workflow & Automation
| Tool |
Description |
| n8n |
AI agent nodes for building LLM-powered automation workflows. |
| Zapier AI |
8,000+ app connections with AI-powered workflows, agents, and chatbots. |
| Make (Integromat) |
Visual automation with detailed branching logic and deep app integrations. |
| Lindy AI |
Agent-first automation for tasks like lead qualification and email follow-ups. |
| Relevance AI |
No-code platform for building collaborating AI agent teams. |
| Activepieces |
MIT-licensed no-code automation with AI workflow integration and self-hosting. |
| Pipedream |
Developer-focused hybrid no-code/serverless platform for API workflows with code at any step. |
| Inngest |
Durable functions replacing queues for reliable multi-step AI workflows. |
| Trigger.dev |
TypeScript-first platform for deploying managed AI agents and background workflows. |
| Windmill |
Workflow engine for internal tools, background jobs, and AI pipelines. |
| Hatchet |
Distributed task queue for reliable background jobs and AI agent workflows. |
| Relay.app |
Human-in-the-loop automation combining AI steps with human approval. |
(⬆ back to top)
🔌 AI Function Calling & Tool Use
| Tool |
Description |
| Composio |
150+ pre-built tool integrations for AI agents with authentication handling. |
| Toolhouse |
Universal tool server enabling any LLM to use tools via a simple API. |
| Arcade AI |
Authenticated API access to third-party services for AI agents. |
| NPI (Natural Programming Interface) |
Tool-use APIs enabling AI agents to operate software tools. |
| Dify |
Platform (93K+ stars) for AI agents with built-in RAG, Function Calling, and ReAct strategies. |
(⬆ back to top)
🧪 AI Testing & QA
| Tool |
Description |
| testRigor |
Generative AI test automation from plain English descriptions. |
| Applitools |
Visual AI detecting UI bugs across browsers and devices. |
| Testsigma |
Natural language test creation for web, mobile, and API. |
| Katalon |
AI-powered test automation for web, mobile, API, and desktop. |
| mabl |
Low-code, AI-driven test automation for Agile and DevOps. |
| QA Wolf |
AI + human-in-the-loop QA with Playwright-based testing and massive parallelization. |
| Qodo |
Generates meaningful test suites and improves code quality. |
(⬆ back to top)
🧬 AI Synthetic Data Generation
| Tool |
Description |
| Gretel |
Fine-tuning capabilities, privacy metrics, and domain-specific generation APIs. |
| Mostly AI |
Transforms production data into privacy-safe versions for AI training. |
| Tonic.ai |
Fabricate (synthetic data), Structural (test data), and Textual (data redaction). |
| YData |
Automated data profiling with synthetic generation for improving AI training quality. |
| Hazy |
Privacy-preserving synthetic data for regulated industries using differential privacy. |
| Synthesis AI |
Specialized in computer vision training with photorealistic 3D-rendered images. |
(⬆ back to top)
🏷️ AI Data Labeling & Annotation
| Tool |
Description |
| Scale AI |
Data labeling with RLHF capabilities, used by major AI labs. |
| Labelbox |
Annotation, data management, and Model Foundry with HIPAA/SOC2 compliance. |
| Snorkel AI |
Programmatic labeling using weak supervision to encode labeling logic as functions. |
| Argilla |
Free data annotation for NLP, LLMs, and multimodal models with Hugging Face integration. |
| Label Studio |
Flexible templates, ML backends, and model-in-the-loop support. |
| Encord |
Computer vision labeling with active learning and model-assisted labeling. |
| CVAT |
Annotation tool for computer vision including image and video labeling. |
(⬆ back to top)
🔗 MCP (Model Context Protocol)
Anthropic's open standard (donated to Linux Foundation AAIF) for connecting AI models to external tools and data sources.
Core Protocol
Notable MCP Servers
| Tool |
Stars |
Description |
| MCP Servers Repository |
91k |
Official collection of reference server implementations. |
| Awesome MCP Servers |
96k |
Community-curated collection covering 7,260+ servers across diverse integrations. |
| Playwright MCP |
38k |
Microsoft's server enabling LLMs to interact with web pages through accessibility snapshots. |
| ActionKit by Paragon |
- |
Connects to 130+ SaaS integrations for AI agent workflows. |
(⬆ back to top)
🛠️ AI Developer Tools & SDKs
| Tool |
Description |
| Vercel AI SDK |
TypeScript toolkit for React/Next.js AI apps with composable agents and MCP support. |
| Streamlit |
Python framework for data-heavy AI apps combining chat interfaces with dashboards. |
| Gradio |
Python library for ML demo UIs with instant public links. |
| Chainlit |
Python framework for conversational AI UIs with step-by-step visualization. |
| Langflow |
Low-code Python-based builder for visually creating agentic and RAG applications. |
| Flowise |
Node.js visual AI workflow builder for agents and chatbots. |
| Dify |
LLM app development platform with visual workflow builder and model management. |
(⬆ back to top)
⏱️ Realtime & Streaming AI
| Tool |
Description |
| OpenAI Realtime API |
Speech-to-speech via WebSocket for real-time voice applications. |
| Google Gemini Live API |
Real-time voice and video interaction with Gemini via WebSocket. |
| LiveKit |
Real-time audio/video AI applications with WebRTC and agent frameworks. |
| Deepgram Streaming |
Real-time STT with sub-250ms latency via WebSocket for live transcription. |
| AG-UI Protocol |
Protocol for streaming agent events between AI backends and frontend UIs. |
| Decart |
Lucy 2 model generating video live with near-zero latency. |
(⬆ back to top)
🎭 Multi-Modal AI APIs
| Tool |
Description |
| OpenAI GPT-4o API |
Natively multimodal: text, images, audio, and video with unified reasoning. |
| Google Gemini API |
1M+ token context with native image/video/audio understanding. |
| Anthropic Claude API |
Vision-capable: analyzes images, documents, charts, and screenshots alongside text. |
| Mistral Pixtral |
Image understanding capabilities via API. |
| Meta Llama Multimodal |
Vision capabilities via various inference providers. |
| Twelve Labs |
Video understanding: search, classification, and generation from video. |
| Roboflow |
Computer vision APIs for training, deploying, and managing detection and segmentation models. |
(⬆ back to top)
📱 Edge & On-Device AI
| Tool |
Stars |
Description |
| Ollama |
- |
Run LLMs locally with a single command; wide model support with easy management. |
| LM Studio |
- |
Desktop app for running LLMs locally with Vulkan GPU offloading and model discovery. |
| llama.cpp |
130k |
C/C++ inference engine optimized for CPU-only inference with extensive quantization. |
| vLLM |
- |
High-throughput serving with PagedAttention and continuous batching. |
| NVIDIA TensorRT-LLM |
15k |
Inference runtime and optimization toolkit for large language models on NVIDIA GPUs. |
| MLX |
29k |
Apple's array and machine-learning framework optimized for Apple silicon. |
| ExecuTorch |
5k |
Meta's on-device framework with 50KB footprint supporting 12+ hardware backends. |
| Jan |
- |
Offline-first desktop AI assistant for running LLMs locally with privacy-first design. |
| LocalAI |
- |
Self-hosted OpenAI-compatible API for running models locally. |
| LlamaEdge |
- |
Lightweight runtime for deploying LLMs at the edge using WebAssembly. |
(⬆ back to top)
🖥️ Cloud GPU Providers
| Tool |
Description |
| RunPod |
On-demand and serverless compute supporting A100/H100 GPUs at flexible pricing. |
| Lambda Cloud |
H100/A100 instances with zero egress fees and InfiniBand networking. |
| CoreWeave |
Kubernetes-native GPU cloud at hyperscale; serves OpenAI and Microsoft. |
| Vast.ai |
GPU marketplace with competitive bidding, best for cost-sensitive workloads. |
| Nebius |
AI-native cloud with Blackwell Ultra compute and European data centers. |
| TensorDock |
GPU marketplace spanning 100+ global locations. |
| Paperspace (DigitalOcean) |
Notebooks, virtual desktops, and serverless functions for ML development. |
| Crusoe Cloud |
Clean-energy-powered GPU cloud for sustainable AI workloads. |
| Fluidstack |
Distributed GPU cloud aggregating idle compute at competitive prices. |
(⬆ back to top)
🧩 AI Browser Extensions & Desktop Apps
| Tool |
Description |
| ChatGPT Desktop |
Native desktop app with system-wide access via global shortcuts. |
| Claude Desktop |
MCP-based extensions and local tool integration. |
| Poe |
Multi-model platform accessing GPT-4, Claude, Gemini, Llama, and custom bots in one interface. |
| Perplexity |
AI search engine as web app and browser extension providing cited answers. |
| Raycast AI |
macOS launcher with built-in AI and extensions for system-wide access. |
| Monica |
Browser extension accessing GPT, Claude, Gemini with sidebar chat. |
| Sider |
Browser sidebar unifying GPT, DeepSeek, Gemini, Claude, and Grok for comparing outputs. |
| Pieces for Developers |
Captures live context from browsers to IDEs with multiple LLM support. |
| HARPA AI |
Chrome extension combining ChatGPT, Claude, Gemini, Perplexity for web page automation. |
(⬆ back to top)
🎮 AI Playgrounds & API Testing
| Tool |
Stars |
Description |
| OpenAI Playground |
- |
Official tool for experimenting with OpenAI models and testing prompts. |
| Vercel AI Playground |
- |
Compare AI model outputs across providers side-by-side. |
| TypingMind |
- |
LLM frontend supporting multiple models with your own API keys and custom plugins. |
| ChatHub |
- |
Browser extension for using ChatGPT, Claude, Gemini side-by-side. |
| Msty |
- |
Desktop app for managing multiple LLM providers with conversation branching and prompt library. |
| OpenPlayground |
6k |
LLM playground for running and comparing models locally. |
(⬆ back to top)
🕷️ Web Scraping & Data Extraction for AI
| Tool |
Stars |
Description |
| Firecrawl |
- |
Scrapes, crawls, and extracts structured data into LLM-ready formats. |
| Crawl4AI |
85k |
AI-ready web crawler generating clean Markdown with local LLM support. |
| Spider |
- |
Fast web scraping designed for AI with structured data extraction. |
| Jina Reader |
- |
Converts any URL into LLM-friendly text. |
| ScrapeGraphAI |
- |
Scraping library using LLMs to create pipelines from natural language. |
(⬆ back to top)
📖 AI Knowledge Management
| Tool |
Description |
| Glean |
Enterprise search connecting Google Workspace, Microsoft 365, Slack, and Salesforce. |
| Notion AI |
AI layer combining wiki, project management, and databases with AI search and writing. |
| Guru |
Knowledge verification delivering knowledge at point-of-work via Slack and Teams. |
| Mem |
Note-taking that self-organizes information using machine learning. |
| Slite |
Collaborative docs using AI to categorize content, identify gaps, and surface relevant info. |
| Confluence AI |
Atlassian's enterprise wiki with AI-powered search and content generation. |
(⬆ back to top)
🏆 Benchmarks & Leaderboards
| Tool |
Description |
| Chatbot Arena (LMArena) |
Crowdsourced LLM ranking using 6M+ blind pairwise votes and Elo ratings. |
| Artificial Analysis |
Independent comparison of 100+ models across quality, price, speed, and latency. |
| Open LLM Leaderboard |
Community benchmark tracking open-source model performance across standardized evaluations. |
| Stanford HELM |
Holistic evaluation across 42 scenarios and 7 dimensions including accuracy, bias, and toxicity. |
| LiveBench |
Continuously updated with new questions to prevent contamination. |
| LLM Stats |
Aggregates multiple leaderboards and benchmarks in one place. |
| MTEB (Massive Text Embedding Benchmark) |
Evaluates embedding models across retrieval, classification, and clustering. |
| EvalPlus |
80x more test cases than HumanEval to address overfitting. |
| BigCodeBench |
1,140 tasks requiring composing function calls from 139 libraries. |
(⬆ back to top)
📚 Learning Resources
Guides & Tutorials
Courses
Books
Newsletters
| Resource |
Description |
| The Batch |
Andrew Ng's weekly AI news for engineers, executives, and enthusiasts. |
| TLDR AI |
Daily AI newsletter covering technical developments for 1.25M+ readers. |
| Ben's Bites |
Daily newsletter with 120K+ subscribers on product launches and business use cases. |
| The Neuron |
Morning-Brew-style AI newsletter for 550K+ professionals. |
| Latent Space |
Technical AI newsletter and podcast on research and infrastructure. |
| Simon Willison's Weblog |
Blog covering LLMs, AI tools, and open source from a prolific builder. |
Podcasts
YouTube Channels
| Resource |
Description |
| Andrej Karpathy |
Former Tesla AI Director teaching neural networks from scratch in "Zero to Hero.". |
| 3Blue1Brown |
Beautifully animated explanations of mathematics, neural networks, and deep learning. |
| Yannic Kilcher |
Detailed breakdowns of the latest ML research papers. |
| AI Explained |
Clear explanations of AI concepts and model comparisons. |
| Two Minute Papers |
Bite-sized summaries of cutting-edge AI research with visual explanations. |
| Umar Jamil |
Highly technical implementations of ML and LLM techniques from scratch. |
Communities
Research Paper Aggregators
| Resource |
Description |
| arXiv |
Open-access preprint repository with 2M+ papers in AI, ML, and CS. |
| Semantic Scholar |
AI-powered tool for finding and connecting scientific papers. |
| Papers With Code |
ML papers linked with code implementations and benchmark results. |
| Connected Papers |
Explores related papers through interactive citation graphs. |
| Hugging Face Daily Papers |
Curated trending AI papers with links to models and datasets. |
Conferences
| Resource |
Description |
| NeurIPS |
Premier ML conference (December annually). |
| ICML |
International Conference on Machine Learning (July 2026, Seoul). |
| ICLR |
International Conference on Learning Representations (April 2026, Rio de Janeiro). |
| AI Engineer World's Fair |
Practitioner-focused conference (June-July 2026, San Francisco). |
(⬆ back to top)
🔗 Related Awesome Lists
(⬆ back to top)
🪦 Deprecated Tools Graveyard
| Dead Tool |
Why |
Use Instead |
| OpenAI Codex |
Shut down Mar 2023 |
GPT-4o, Claude Code. |
| Hugging Face Inference API (free) |
Removed free tier |
Hugging Face Inference Endpoints (paid). |
| Jasper Boss Mode |
Rebranded/discontinued |
Jasper AI platform. |
| ChatGPT Plugins |
Shut down Apr 2024 |
GPTs, function calling. |
| Bard |
Renamed to Gemini |
Google Gemini. |
| LangChain Agents (legacy) |
Deprecated in favor of LangGraph |
LangGraph. |
(⬆ back to top)
Contributing
Contributions welcome! Please read the contribution guidelines first.
- Add tools you've actually used or evaluated
- Include star count and language where applicable
- Note if a tool is unmaintained (last commit >1 year ago)
- Commercial tools/services are fine but must be clearly labeled
To the extent possible under law, Edward Tay has waived all copyright and related or neighboring rights to this work.