Skip to content
The AI Landscape

The AI landscape is vast, and it's easy to get lost in.

New models, new tools, new names land constantly. Some of it is noise. Some of it changes what we'd recommend to you. We keep track of which is which.

The six layers

Strip away the noise and it's six layers.

01 Model providers
02 Open models
03 AI apps & agents
04 Media generation
05 Agent building
06 Infrastructure & local AI

Open any layer.

The companies we use, deploy, and track for clients. What each one makes, and where it sits. These are the biggest names in each layer, not every name.

There's no official map of AI. Chipmakers draw it as a hardware stack, investors draw market maps, and vendors draw it around whatever they sell. Six layers is how we explain it to clients: each layer is a different kind of decision, from which model to build on to where your data lives.

Nothing here is sponsored. Last reviewed July 2026.

OpenAI

Builds general-purpose models for chat, coding, and images, for consumers and via API.

Anthropic Anthropic

Builds the Claude model family, with products for chat, autonomous work, coding, and API access.

Google Google

Builds multimodal models and AI tools across consumer apps, developer APIs, and research.

xAI

Builds the Grok model family for chat, image and video generation, voice, and developer APIs.

Meta Meta

Builds Meta AI and the Muse frontier models; keeps open tools like PyTorch and SAM, plus the earlier open Llama family.

Cohere

Builds enterprise models and a secure agent platform, with a retrieval and multilingual focus.

Mistral AI Mistral AI

European lab that releases open-weight models alongside hosted chat, coding, and developer platforms.

DeepSeek DeepSeek

Chinese lab releasing open-weight language and reasoning models, with a free chat app and API.

QWen Qwen

Alibaba's open-weight model family spanning text, multimodal, and coding, with a chat platform and API.

Moonshot AI Moonshot AI

Chinese lab releasing the open-weight Kimi models, with a chat app, agents, and a developer API.

Z.ai Z.ai

Chinese lab releasing the open-weight GLM model family for chat, coding, and reasoning, with a chat app and API.

TII TII

Abu Dhabi's Technology Innovation Institute, releasing the open-weight Falcon models, including Arabic-language models.

Microsoft

Copilot assistants across Microsoft 365, GitHub, and Windows, plus its own MAI models, built on Azure.

Perplexity Perplexity

AI answer engine with a web browser, a desktop agent, and a developer search API.

Cursor Cursor

AI code editor with autonomous agents, inline completion, and automated code review.

Notion Notion

Connected workspace with built-in AI agents that draft, search, and run work across your docs and tools.

OpenClaw OpenClaw

Open-source autonomous agent operated through chat, running tasks on your own machine.

Cognition

Maker of Devin, the autonomous software engineer, spanning cloud agents and the Devin Desktop editor.

Manus Manus

Autonomous agent that plans and completes multi-step research and web tasks on its own.

Replit Replit

Browser platform that turns prompts into deployed apps, with hosting built in.

Lovable

AI builder that generates and hosts full web apps from plain-language prompts.

Midjourney logo Midjourney

Text-to-image and image-to-video generation, used through its web app and Discord.

Runway Runway

AI video generation, in-context video editing, and world models for film and production work.

ByteDance ByteDance

TikTok's parent company; builds the Seedance video and Seedream image models and the Doubao assistant.

Black Forest Labs Black Forest Labs

Image-generation models offered both as a hosted API and as open weights for self-hosting.

ElevenLabs ElevenLabs

Voice synthesis, conversational voice agents, music, and audio tools, via apps and an API.

HeyGen

AI avatar video, multilingual video translation, and an API for programmatic video.

Higgsfield

AI image and video generation with camera-motion control, plus lip-sync and marketing tools.

Kling AI

Text- and image-to-video generation with synced audio, built by the Chinese company Kuaishou.

MiniMax MiniMax

Chinese lab behind the Hailuo video platform, alongside open-weight language models and a developer API.

Luma AI

Video generation and creative AI agents; its Ray models turn text and images into video.

Suno Suno

AI music generation. Full songs, vocals included, from a text prompt, via web and mobile apps.

n8n n8n

Visual workflow automation for building AI agents and integrations, self-hostable or cloud-hosted.

Zapier Zapier

No-code automation across thousands of apps, with AI agents and an MCP bridge to the major assistants.

Make Make

Visual automation platform for connecting apps and building AI agents into everyday business workflows.

LangChain LangChain

Open-source frameworks for building LLM agents, plus a hosted platform for tracing, evals, and deployment.

LlamaIndex LlamaIndex

Open-source framework and hosted platform for document parsing and retrieval over complex data.

CrewAI CrewAI

Open-source framework for orchestrating multi-agent systems, with a hosted management platform.

Pydantic Pydantic AI

Type-safe Python agent framework from the Pydantic team, paired with the Logfire observability platform.

NVIDIA NVIDIA

Supplies the chips, software platforms, and open models that AI systems are built and run on.

AWS

Cloud platform with managed model APIs, its own Nova models, ML infrastructure, and custom AI chips.

Hugging Face Hugging Face

Repository and platform for open models, datasets, and demos, with hosted inference.

OpenRouter OpenRouter

One API in front of hundreds of models from every major lab, with routing, fallbacks, and unified billing.

Groq Groq

Inference provider running open-weight models on its own LPU hardware, in the cloud or on-prem.

Cerebras Cerebras

Builds wafer-scale AI chips and a fast inference cloud, an alternative to GPU-based compute.

together.ai Together AI

Cloud platform for running, fine-tuning, and hosting open-weight models on managed GPUs.

Replicate Replicate

Cloud platform for running open-weight models via API, with an open-source packaging tool.

Vercel Vercel

Web infrastructure for apps and agents, with a TypeScript AI SDK, a multi-provider model gateway, and AI dev tools.

Ollama Ollama

Runs open-weight models locally via a CLI and local API, with an optional hosted service.

LM Studio LM Studio

Desktop app for discovering and running local models, with a headless server, CLI, and SDKs.

Where we come in

You don't need to know all of this. You need someone who does.

That's the job. We track the landscape so you can make one good decision at a time. Book a call and tell us what you're weighing.