The canonical catalog of LLMs, frameworks, vector stores, and infrastructure powering modern AI products. Click any tech to see every startup in our directory that runs on it.
79 technologies tracked across 10 categories.
OpenAI's flagship multimodal reasoning model. The category default for most production AI apps.
OpenAI's omni model — cheap, fast, and good enough for most tasks.
OpenAI's lightweight workhorse. The de facto default for cheap-tier inference.
Anthropic's balanced model — the developer's favorite for code and long-context work.
Anthropic's flagship frontier model, tuned for nuanced reasoning.
Anthropic's fastest, cheapest tier — ideal for classification and routing.
Google's long-context flagship — 1M+ token window for document-heavy workloads.
Google's cheap-tier Gemini variant — fast and economical.
Meta's open-weight foundation model family, widely fine-tuned.
European frontier LLM from the Paris lab.
Open-weight mixture-of-experts model — strong performance, permissive license.
Chinese frontier lab delivering extremely cost-efficient reasoning models.
Alibaba's open-weight LLM family, strong on multilingual and code tasks.
xAI's flagship model, tightly integrated with the X platform.
Cohere's enterprise-grade retrieval-tuned LLM family.
Unified API gateway across hundreds of hosted LLMs.
Ultra-low-latency LPU inference for open-weight models.
Hosted inference + fine-tuning for hundreds of open-weight models.
Web-grounded LLM API — Perplexity's search-augmented chat endpoint.
The most widely used LLM orchestration framework. Python and JS.
Data framework for LLM apps — indexing, retrieval, and agents.
TypeScript-first toolkit for streaming UIs and multi-provider LLM calls.
Open-source Python framework for production LLM and search pipelines.
Declarative prompt programming — compile prompts, not hand-craft them.
Structured-output wrapper for LLMs — Pydantic in, typed data out.
React framework powering the majority of AI startup frontends.
Python web framework of choice for AI backend APIs.
Fully managed vector database — the incumbent in production RAG.
Open-source vector database with built-in hybrid search.
Open-source vector search engine in Rust — fast and lean.
Open-source embeddings database optimized for prototypes and local dev.
Cloud-native open-source vector database built for billion-scale.
Vector similarity search inside Postgres — the pragmatic default.
Managed pgvector on Supabase — embeddings without leaving Postgres.
Serverless vector database built on Apache Arrow.
Run and fine-tune open-source models in the cloud with one API.
Frontend cloud — Next.js-native deployments with edge functions.
Edge serverless runtime on Cloudflare's global network.
AWS managed foundation-model service — multi-vendor LLM hosting.
Amazon Web Services — the default public cloud.
Google's public cloud — strong on AI/ML and data.
Azure's enterprise-grade hosting of OpenAI models.
Serverless compute for AI — GPU in a Python decorator.
Affordable GPU cloud for training and inference.
GPU cloud for AI training — H100/A100 instances at scale.
Specialized GPU cloud powering frontier model training.
Open-source Firebase alternative — Postgres + auth + storage.
Serverless Postgres with branching for AI app backends.
The default hub for models, datasets, and hosted inference endpoints.
Deploy apps close to users with Firecracker microVMs.
Trace + evaluate LLM apps — from the LangChain team.
LLM observability and caching proxy.
Open-source LLM engineering platform — traces, evals, prompts.
End-to-end eval platform for AI products.
ML experiment tracking and model evaluation for AI teams.
LLM and ML observability platform focused on drift and quality.
Open-weight image model family — the base for countless products.
OpenAI's image generation model, accessible via API and ChatGPT.
Open-weight flow-matching image model — current SOTA for quality.
Category-leading diffusion model known for distinctive aesthetic.
Generative video and image tooling for creators.
Turn websites into LLM-ready markdown with one API call.
Independent search index with a developer-friendly API.
Fast Google Search API for AI agents.
Search API purpose-built for LLM agents and RAG.
Semantic web search API designed for AI applications.
Default payments API for AI startups monetizing via subscriptions.
Drop-in auth and user management for modern apps.
Enterprise-grade identity and authentication platform.