A map of the AI tools worth building with.

Langfuse

Observability & evals

An open-source platform for tracing, evaluating, and monitoring LLM and agent applications.

LangGraph

Agent frameworks

A low-level framework for building stateful, multi-step agents as explicit graphs of nodes and edges.

LiteLLM

Gateways & routers

An open-source proxy that gives 100+ model providers one OpenAI-compatible API.

Llama

Open-weight models

Meta's family of open-weight models, the most widely deployed base for self-hosted AI.

LlamaIndex

RAG & knowledge

A data framework for connecting language models to your own documents and structured data.

Model Context Protocol

MCP & connectors

An open standard for connecting AI models to tools and data through a common protocol.

n8n

Workflow & automation

A fair-code workflow automation tool with 400+ integrations and AI nodes, self-hostable end to end.

Ollama

Model serving

The simplest way to download and run open-weight models locally, with one command.

Open WebUI

Chat & assistant UIs

A self-hosted, feature-rich chat interface for local and remote models, the most common front door for a private assistant.

OpenAI

LLM providers

The most widely adopted frontier-model provider, the usual default for a new AI project.

pgvector

Vector databases

A PostgreSQL extension that adds vector similarity search to a database you already run.

Playwright

Browser automation

Microsoft's modern browser automation library, the reliable base for driving a real browser.

Unstructured

Document extraction

A toolkit and API for turning PDFs, office files, and emails into clean, chunked text for AI.

Vercel

Hosting & infrastructure

A developer platform for deploying web apps and AI front-ends with minimal configuration.

Whisper

Voice & transcription

OpenAI's open-source speech-to-text model, the default starting point for transcription.

Activepieces

Workflow & automation

An open-source, no-code automation tool positioned as a self-hostable Zapier alternative.

Agno

Agent frameworks

A fast, lightweight Python framework for building multi-agent systems with memory and tools.

Amazon Bedrock

LLM providers

AWS's managed service for calling many foundation models behind one API.

Amazon Web Services

Hosting & infrastructure

The broadest cloud platform: compute, storage, GPUs, and managed AI services at any scale.

Anthropic

LLM providers

The maker of Claude, frontier models known for strong reasoning, long context, and tool use.

AnythingLLM

Chat & assistant UIs

An all-in-one desktop or self-hosted app that turns a folder of documents into a chat assistant with workspaces.

Apache Airflow

Workflow & automation

The long-standing open-source platform for authoring, scheduling, and monitoring data pipelines as code.

Arcade

MCP & connectors

A platform for giving agents authenticated tool access, handling OAuth and per-user permissions.

AssemblyAI

Voice & transcription

A commercial speech-to-text API pairing transcription with audio-intelligence models.

AutoGen

Agent frameworks

Microsoft's framework for multi-agent systems built around asynchronous, conversational message passing.

Azure AI Foundry

LLM providers

Microsoft's platform for accessing OpenAI and other models within the Azure cloud.

Baseten

Model serving

A platform for deploying and serving custom and open models on autoscaling infrastructure.

Bifrost

Gateways & routers

A fast, open-source LLM gateway written in Go, with one API across providers.

big-AGI

Chat & assistant UIs

A feature-rich open-source AI suite with personas, multi-model chat, and side-by-side response comparison.

Braintrust

Observability & evals

A commercial platform for evaluating, logging, and iterating on AI products.

Browser Use

Browser automation

An open-source library that lets an AI agent control a browser from natural-language goals.

Browserbase

Browser automation

A hosted, scalable headless-browser platform purpose-built for AI agents that browse the web.

Browserless

Browser automation

A self-hostable service for running headless browsers behind a managed API.

Cartesia

Voice & transcription

A commercial provider of very low-latency, realistic text-to-speech for real-time voice apps.

Chatbox

Chat & assistant UIs

A cross-platform desktop client for chatting with many model providers and local models from one app.

Chroma

Vector databases

An open-source embedding database designed to be the simplest way to start with RAG.

Claude Agent SDK

Agent frameworks

Anthropic's SDK for building agents on Claude, with tools, subagents, and managed context.

Cloudflare AI Gateway

Gateways & routers

Cloudflare's proxy for model traffic, adding caching, rate limiting, and analytics at the edge.

Cloudflare Workers

Hosting & infrastructure

A serverless platform that runs code at the edge, with AI, storage, and database primitives.

Cognita

RAG & knowledge

An open-source, modular RAG framework from TrueFoundry for taking pipelines to production.

Cohere

LLM providers

An enterprise-focused AI provider with strong models for retrieval, ranking, and generation.

Command R

Open-weight models

Cohere's open-weight models tuned for retrieval-augmented generation and tool use.

Composio

MCP & connectors

A platform that gives agents authenticated access to hundreds of apps through managed tools.

Coolify

Hosting & infrastructure

An open-source, self-hostable platform for deploying apps, a Heroku and Netlify alternative.

CrewAI

Agent frameworks

A framework for orchestrating role-based agents that collaborate on a task as a crew.

DeepEval

Observability & evals

An open-source "Pytest for LLMs", a unit-testing framework for model outputs.

Deepgram

Voice & transcription

A commercial speech API built for fast, accurate transcription at scale, including real-time.

DeepSeek

Open-weight models

A family of open-weight models known for strong reasoning and coding at a low cost to run.

Dify

Workflow & automation

An open-source platform for building LLM apps and agentic workflows on a visual canvas.

Docling

Document extraction

IBM's open-source library for parsing documents into a structured, AI-ready representation.

DSPy

Agent frameworks

A framework for programming language models with optimisable modules instead of hand-written prompts.

Elasticsearch

Vector databases

The widely deployed search engine, now with dense vector and hybrid search built in.

ElevenLabs

Voice & transcription

A leading text-to-speech and voice platform known for highly natural, expressive synthetic speech.

faster-whisper

Voice & transcription

A fast reimplementation of Whisper on CTranslate2, with much lower latency and memory use.

FastMCP

MCP & connectors

The most-used framework for building MCP servers in Python, with a fast, decorator-based API.

Fireworks AI

Model serving

A hosted inference platform focused on fast, low-cost serving of open models.

Flowise

Workflow & automation

A visual drag-and-drop builder for LLM flows and agents, open source and self-hostable.

Fly.io

Hosting & infrastructure

A platform for running containers close to users, with easy global and GPU deployment.

Gemma

Open-weight models

Google's family of small, capable open-weight models built from Gemini research.

Google Cloud

Hosting & infrastructure

Google Cloud: compute, data, and AI infrastructure, including Vertex AI and Gemini.

Google Gemini

LLM providers

Google's Gemini models, offering very long context and strong native multimodal support.

Groq

Model serving

A hosted inference service running open models on custom hardware for very low latency.

Haystack

RAG & knowledge

deepset's open-source framework for composable, production-grade search and RAG pipelines.

Helicone

Observability & evals

An open-source observability platform that logs and analyses LLM calls through a proxy.

A minimal, self-hostable web UI for Ollama and OpenAI-compatible endpoints, with no database to run.

Inngest

Workflow & automation

A durable workflow engine for developers, with event-driven steps, retries, and flow control as code.

Jan

Chat & assistant UIs

An open-source, offline-first desktop app for running AI models locally, a private ChatGPT alternative.

Klavis AI

MCP & connectors

An open-source project for hosted, production-ready MCP servers with built-in authentication.

A small, open-weight text-to-speech model that produces natural voices on modest hardware.

LanceDB

Vector databases

An open-source, embedded vector database built on the Lance columnar format.

Langflow

Workflow & automation

An open-source visual builder for LLM pipelines and agents, Python-based.

LangSmith

Observability & evals

LangChain's platform for tracing, testing, and evaluating LLM and agent applications.

LibreChat

Chat & assistant UIs

An open-source ChatGPT-style interface that unifies many model providers behind one familiar, extensible UI.

llama.cpp

Model serving

An open-source C/C++ runtime for running LLMs efficiently on CPUs and consumer GPUs.

LlamaParse

Document extraction

A hosted parsing service tuned for extracting clean structure from complex, messy PDFs.

LLMWare

RAG & knowledge

An open-source framework for enterprise RAG pipelines tuned around small, specialised models.

LM Studio

Model serving

A desktop app for discovering, downloading, and running local models with a graphical UI.

LobeChat

Chat & assistant UIs

A polished open-source chat UI with a plugin system, custom agents, and broad multi-model support.

An open-source, OpenAI-compatible API you can run locally over many model backends.

Lunary

Observability & evals

An open-source observability and prompt-management platform for LLM applications.

Make

Workflow & automation

A visual automation platform for connecting apps into multi-step scenarios, with a more flexible canvas than Zapier.

Marker

Document extraction

An open-source tool that converts PDFs, ePub, and more into clean Markdown quickly and accurately.

Mastra

Agent frameworks

A TypeScript framework for agents, workflows, and RAG, from the team behind Gatsby.

mcp-use

MCP & connectors

An open-source client library for connecting any LLM to MCP servers in a few lines of code.

Microsoft Azure

Hosting & infrastructure

Microsoft's cloud platform, with deep integration for OpenAI models and enterprise workloads.

Milvus

Vector databases

An open-source vector database built for billion-scale similarity search.

MinerU

Document extraction

An open-source tool that extracts structured content from PDFs, preserving layout, tables, and formulas.

Mistral (open models)

Open-weight models

Mistral's family of open-weight models, from small efficient ones to mixture-of-experts.

Mistral AI

LLM providers

A European AI lab offering frontier and efficient models through a developer API.

Modal

Model serving

A serverless platform for running Python (including model inference) on cloud GPUs.

A desktop app for local and remote models, focused on a frictionless one-click setup.

OLMo

Open-weight models

Allen AI's truly open language models, released with full data, code, and training logs.

OpenAI Agents SDK

Agent frameworks

OpenAI's lightweight framework for building agents with handoffs, guardrails, and built-in tracing.

OpenLLMetry

Observability & evals

An open-source set of OpenTelemetry extensions for standardised LLM observability.

OpenRouter

Gateways & routers

A hosted marketplace and router that exposes hundreds of models behind one API and one bill.

Opik

Observability & evals

Comet's open-source platform for tracing, evaluating, and monitoring LLM applications.

PaddleOCR

Document extraction

A widely used open-source OCR toolkit supporting many languages and document layouts.

Phi

Open-weight models

Microsoft's Phi family of small models trained for strong reasoning at low cost.

Phoenix

Observability & evals

Arize's source-available tool for tracing, evaluating, and debugging LLM and agent apps.

Pinecone

Vector databases

A fully managed vector database that removes all operational work from similarity search.

Pipedream

MCP & connectors

A connector platform with thousands of integrations, now exposing tools to agents over MCP.

Piper

Voice & transcription

A fast, local open-source text-to-speech system designed to run well on small devices.

Playwright MCP

Browser automation

An MCP server that exposes Playwright browser control to any MCP-aware model or agent.

Portkey

Gateways & routers

An AI gateway pairing fast model routing with built-in observability, caching, and guardrails.

Prefect

Workflow & automation

A Python-native orchestration tool for building, running, and observing data and AI workflows.

PrivateGPT

RAG & knowledge

An open-source project for asking questions of your documents fully offline and private.

promptfoo

Observability & evals

An open-source tool for testing, evaluating, and red-teaming prompts and models.

Puppeteer

Browser automation

A mature open-source library for driving headless Chrome and Firefox from Node.js.

Pydantic AI

Agent frameworks

A type-safe Python agent framework from the Pydantic team, built around validated structured outputs.

Qdrant

Vector databases

A fast open-source vector database with rich filtering, built for production retrieval.

An open-source RAG framework and "second brain" for chatting with your files and data.

Qwen

Open-weight models

Alibaba's open-weight model family, spanning many sizes with strong multilingual coverage.

R2R

RAG & knowledge

An open-source "RAG to riches" engine with ingestion, retrieval, and agentic search built in.

Ragas

Observability & evals

An open-source framework for evaluating RAG pipelines on faithfulness and relevance.

RAGFlow

RAG & knowledge

An open-source RAG engine built on deep document understanding, designed to run as a complete app.

Railway

Hosting & infrastructure

A developer platform for deploying apps, databases, and services with minimal configuration.

Redis

Vector databases

Vector search inside Redis: similarity search alongside the cache you already run.

Reducto

Document extraction

A commercial document-ingestion API focused on high-accuracy parsing of complex, regulated documents.

Render

Hosting & infrastructure

A managed cloud for deploying web services, static sites, background workers, and databases.

Replicate

Model serving

A platform for running and fine-tuning open models behind a simple hosted API.

Requesty

Gateways & routers

A routing layer that unifies many models with spend controls and automatic failover.

Selenium

Browser automation

The long-established open-source standard for cross-browser automation and testing.

Semantic Kernel

Agent frameworks

Microsoft's SDK for composing AI skills, plugins, and agents into enterprise applications.

SGLang

Model serving

A fast open-source serving runtime with structured generation and high throughput.

Smithery

MCP & connectors

A registry and hosting platform for discovering and deploying MCP servers.

smolagents

Agent frameworks

Hugging Face's minimal agent library where agents act by writing and running Python code.

SmolLM

Open-weight models

Hugging Face's fully open small-model family, released with open data and training recipes.

Stagehand

Browser automation

An open-source framework that adds natural-language actions on top of Playwright.

Steel

Browser automation

An open-source browser API purpose-built for AI agents that need to use the web.

Supabase

Hosting & infrastructure

An open-source backend platform built on Postgres: database, auth, and storage in one.

Surya

Document extraction

An open-source OCR and layout-analysis toolkit covering text, tables, and reading order in many languages.

Temporal

Workflow & automation

A durable execution engine that makes long-running, failure-prone workflows reliable by default.

Text Generation Inference

Model serving

Hugging Face's production server for high-performance LLM serving.

Together AI

Model serving

A cloud for running and fine-tuning open models with fast, hosted inference.

Trigger.dev

Workflow & automation

An open-source framework for long-running background jobs written as plain TypeScript code.

turbopuffer

Vector databases

A serverless vector and full-text search engine built on object storage for low cost.

txtai

RAG & knowledge

An all-in-one open-source embeddings database for semantic search, RAG, and LLM workflows.

TypingMind

Chat & assistant UIs

A commercial front-end that puts your own provider API keys behind one polished chat UI.

Vapi

Voice & transcription

A platform for building, testing, and deploying real-time voice agents over the phone.

Verba

RAG & knowledge

Weaviate's open-source RAG application, a ready-to-run chatbot for your own documents.

Vercel AI Gateway

Gateways & routers

Vercel's gateway for routing to many model providers through one endpoint and one bill.

Vespa

Vector databases

An open-source engine for large-scale search, recommendation, and vector workloads.

vLLM

Model serving

A high-throughput inference engine for serving open models efficiently in production.

Weaviate

Vector databases

An open-source vector database with hybrid search and built-in vectorisation modules.

An open-source extension of Whisper adding accurate word-level timestamps and speaker diarisation.

Windmill

Workflow & automation

An open-source platform that turns scripts into workflows, internal UIs, and scheduled jobs.

xAI Grok

LLM providers

xAI's Grok family of frontier models, available through a developer API.

Zapier

Workflow & automation

The most widely used hosted automation platform, connecting thousands of apps with no code.