Updated
What makers build

AI coding agents & assistants tech stack 2026: what 63 real products use

Tools that write, edit or explain code in an editor, terminal or browser.

Among the 63 products we track in AI coding agents & assistants, the usual stack is React (frontend framework, 69%), Vercel (hosting, 44%), Claude (LLM API, 76%) and PostgreSQL (database, 46%): shares of the products that show each decision.

How they're built

The typical stack

The leading pick where at least 8 of them show the decision and the leader has at least a quarter of it.

DecisionMost common pickShareAt default usage
Frontend FrameworkReact33 of 48 · 69%—
HostingVercel18 of 41 · 44%$40/mo · Pro + Flat Rate CDN 10M
LLM APIClaude26 of 34 · 76%$10/mo · Claude Haiku 5.5
DatabasePostgreSQL11 of 24 · 46%—
MCP Server for Your ProductMCP SDK24 of 24 · 100%—
AI SDK & Agent FrameworkAI SDK11 of 23 · 48%—
Backend FrameworkExpress8 of 22 · 36%—
AI UI Designshadcn/ui17 of 17 · 100%—
Transactional EmailResend15 of 17 · 88%$20/mo · Pro 50k
Database Access & ORMsDrizzle ORM5 of 12 · 42%—
Product AnalyticsPostHog8 of 11 · 73%$0/mo · Pay-as-you-go
Web AnalyticsVercel Web Analytics5 of 11 · 45%—
AuthenticationClerk4 of 9 · 44%$0/mo · Hobby
Error MonitoringSentry8 of 9 · 89%$29/mo · Team
PaymentsStripe6 of 8 · 75%fees on revenue · $410 on $10,000/mo

Monthly bills add up to about $99 at the calculators' default usage, list prices; fees on revenue (Stripe) come on top. Set your own usage →

Where they differ from other products

Tools AI coding agents & assistants pick at least twice as often as products overall. Each bar is the share of products that show the decision and chose the tool.

What they chose, decision by decision

Among the AI coding agents & assistants that show each choice, from makers' products and open-source code alike. Under each tool, some of the products that chose it.

React33 · 69%Incerto, supercli, omnigent
Next.js27 · 56%RightNow AI, Git Pitcher, stagewise
React Router9 · 19%omnigent, gptme, Archon
Astro4 · 8%Deep Work Plan, Archon, openclaude
Hosting41 products
Vercel18 · 44%RightNow AI, stagewise, OpenCode Superapp
GitHub Pages12 · 29%tabby, atomic, kandev
Cloudflare11 · 27%Deep Work Plan, agent-manager, Termi Protocol
LLM API34 products
Claude26 · 76%RightNow AI, Fixa.dev, Codebuff
OpenAI API22 · 65%RightNow AI, OpenCode Superapp, supercli
Gemini API13 · 38%Incerto, supercli, impeccable
Amazon Bedrock7 · 21%atomic, costrict, openclaude
Database24 products
PostgreSQL11 · 46%Incerto, supercli, omnigent
SQLite7 · 29%omnigent, emdash, Wegent
MCP SDK24 · 100%supercli, omnigent, paseo
FastMCP1 · 4%ai-dev-kit
AI SDK11 · 48%supercli, omnigent, paseo
Claude Agent SDK9 · 39%omnigent, paseo, cyrus
LlamaIndex2 · 9%Wegent, avante.nvim
Pydantic AI2 · 9%code-graph-rag, code_puppy
Express8 · 36%supercli, paseo, cyrus
Hono5 · 23%supercli, kandev, Archon
FastAPI5 · 23%omnigent, AiTerm, Wegent
Flask4 · 18%gptme, ai-dev-kit, kiteco-public
AI UI Design17 products
shadcn/ui17 · 100%supercli, omnigent, compozy
Drizzle ORM5 · 42%emdash, costrict, MiMo-Code
SQLAlchemy4 · 33%omnigent, Wegent, ai-dev-kit
Prisma2 · 17%supercli, repomind
PostHog8 · 73%Git Pitcher, tabby, Archon
Mixpanel1 · 9%kiteco-public
Web Analytics11 products
Vercel Web Analytics5 · 45%supercli, compozy, tabby
Plausible Analytics3 · 27%omnigent, paseo, atomic
Authentication9 products
Clerk4 · 44%Git Pitcher, Fixa.dev, cto.new
Better Auth2 · 22%supercli, Archon
WorkOS2 · 22%packmind, agent-orchestrator
Sentry8 · 89%RightNow AI, Git Pitcher, cyrus
Rollbar1 · 11%kiteco-public

Small samples, fewer than 8 products: Vector Database (Qdrant 3, Milvus 2) · Image & Video Hosting (sharp 6) · Document Parsing (pypdf 4, pdf-parse 1) · File Storage (Amazon S3 2, MinIO 2) · AI Code Review (CodeRabbit 4, Greptile 1) · Browser Automation for Agents (Puppeteer 3, Browserbase 1) · Context & Skills for Coding Agents (Chrome DevTools MCP 2, DeepWiki 1) · LLM Observability & Evals (Langfuse 1, Helicone 1) · Open Model Hosting & GPUs (Together AI 1, Fireworks AI 1) · Background Jobs & Cron (Inngest 1, BullMQ 1) · Mobile Framework (Expo 3, React Native 3) · Parallel Agents & Spec-Driven Work (Beads 2, Conductor 1) · Run Models Locally (Ollama 3)

AI coding agents & assistants we track

35 makers' products and 28 open-source projects, makers' products first, each with the tools it uses (the most common here first).

Questions

What tech stack is behind AI coding agents & assistants?

Among the 63 products we track in AI coding agents & assistants, the usual stack is React (frontend framework, 69%), Vercel (hosting, 44%), Claude (LLM API, 76%) and PostgreSQL (database, 46%): shares of the products that show each decision. After those, MCP SDK leads MCP server for your product, AI SDK leads AI SDK & agent framework and Express leads backend framework.

Which tools are more common in AI coding agents & assistants than elsewhere?

Amazon Bedrock for LLM API (6.1× as often as among products overall), GitHub Pages for hosting (3.9× as often as among products overall) and Claude Agent SDK for AI SDK & agent framework (3.6× as often as among products overall).

How much does a typical AI coding agents & assistants stack cost?

About $99 a month in bills at the pricing calculators' default usage and list prices, plus Stripe fees on revenue. The biggest lines: Vercel $40, Sentry $29 and Resend $20.

Which products in AI coding agents & assistants are open source?

28 of the 63 are open-source projects, known from their public code, among them costrict, openclaude, specweave, codeg, Wegent and MiMo-Code.