AI
Lessons from Building an AI App Builder on Convex
Over the past few months, we have built and grown Chef into the only AI app builder that knows backend. But, this process didn’t start with Chef. It started with building the Convex, the database that Chef is built on top of.
AI Agents (and humans) do better with good abstractions
Chef by Convex builds real full-stack apps in one prompt—Notion, Slack, and more. It works because Convex’s abstractions are simple enough for both humans and AI to use. Built-in features and plug-and-play components let developers skip boilerplate and ship fast.
Claude 4 is here but is is good at Convex?
Claude 4 is here, and developers are putting it to the test. This video skips the benchmarks and dives into real-world builds—like an Instagram clone and a multiplayer Tic-Tac-Toe app—using Claude 4 with Convex and Chef. If you’re deciding between Claude 4 and 3.5 for agentic codegen, backend setup, or Convex-based workflows, this walkthrough shows the good, the bad, and the frustrating. From schema generation to real-time uploads and env var debugging, you’ll see where Claude 4 outperforms—and where Claude 3.5 still holds its ground. Ideal for devs building modern fullstack apps who want to see Claude in action, not just theory.
Chef Cookbook: Tips For Working with AI App Builders
Learn how to write better apps using AI app builders like Chef by Convex with these five expert tips from Jordan Hunt, prompt engineer at Convex. This guide covers how to build simple MVPs, keep prompts under 300 words, provide clear UI and design instructions, use AI tools like ChatGPT to refine your prompts, and recover quickly when things go off track. Includes real app examples like a habit tracker, to-do list, and finance tracker—all built with Chef. Perfect for developers building full-stack apps with AI.
Which LLM writes the best code? Convex Chef model comparison
Convex compared Claude 3.5 Sonnet, GPT-4.1, and Gemini 2.5 Pro to see which LLM works best for building full-stack apps with Convex Chef, the new AI Agent app developer tool. Claude stood out for its precise backend coding and function calling, while Gemini made impressive UIs, and GPT offered solid speed and value.
Agents Need Durable Workflows and Strong Guarantees
Agents rely on long-lived workflows, but when happens when they fail midway through? Here are the tools you need to manage correctness and reliability: transactions, idempotency, retries, durable functions, journaling, and state machines. The missing abstraction layer for agentic is durable workflows, which bring them all together.
AI Agents with Built-in Memory
With this new backend component, augment Agents to automatically save and search message history per-thread, providing realtime results across multiple tabs and users. Use it with the Workflow component to run reliably with retries and durability across server restarts.
Reimplementing Mastra Workflows: Lessons Learned
I reimplemented Mastra’s agentic workflows with durable functions in Convex, and it was the wrong decision. Look at three common strategies (reimplementation, API wrapping, and “blessed” plugin paths), along with learnings along the way and reflections on what I’d do differently next time. TL;DR: Do less, do it smarter, and prototype faster.
Convex Evals: Behind the scenes of AI coding with Convex
AI coding is here: The most productive developers are leveraging AI to speed up their workflows. This ranges from asking models questions about system design to letting AI take the driver's seat with tools like Cursor Composer.
Convex now supports a powerful MCP server that lets you introspect your deployment's state, run functions, and read and write data.
Coding agents can do more autonomously when they write code that has tight, automatic feedback loops; use systems that express everything in standard, procedural code; and have access to strong, foolproof abstractions.
Streaming HTTP Responses using fetch
Learn the basics of HTTP streaming with Convex by re-implementing OpenAI's SDK using built-in fetch and async iterators. No npm dependencies needed.
By leveraging HTTP actions with streaming, this chat app balances real-time responsiveness with efficient bandwidth usage. Users receive character-by-character updates to their own responses directly from ChatGPT, while other users see periodic updates, minimizing database bandwidth.
How I built NotesGPT – a full-stack AI voice note taking app
I recently built a full-stack app called notesGPT. It allows you to record a voice note, transcribes it, and extract action items and display them as action items. It’s fully open source and comes equipped with authentication, storage, vector search, action items, and is fully responsive on mobile for ease of use.
4 Devs, 1 Idea, 4 Apps in 4 Hours(!!) with Convex
Using Convex, 4 web devs built their own fullstack app based on this prompt: > Build a way to show real-time updates on the website for a Dungeons and Dragons-themed small business! See what they built, learn how they did it, and watch their reactions to each other's work in this installment of the "4 Web Devs, 1 App Idea" video series.
This year vector databases have sprung up like mushrooms to enable applications to retrieve context based on semantic search. A large portion of these applications have used the retrieved context to augment the ability of large language models (LLMs) in a pattern known as RAG. On November 7th OpenAI released its Assistants API, enabling the implementation of AI chat interfaces with context retrieval without needing a separate message store or vector database. Does this new API make vector databases obsolete?
Build AI Chat with Convex Vector Search
Convex is a full-stack development platform and cloud database, including built-in vector search. In this third post in our [series](https://stack.convex.dev/ai-chat-using-openai-assistants-api), we’ll build an AI-powered chat interface using Convex, with our own message storage and context retrieval.
Build AI Chat with OpenAI's Assistants API
On November 7th OpenAI released its Assistants API, enabling chat bot with context retrieval implementations without needing a messages or vector database. In this post, we’ll cover how to leverage this API to build a fully functioning AI chat interface.
GPT Streaming With Persistent Reactivity
Stream GPT responses without brittle browser-based HTTP streaming. Multiplayer reactivity, persistence, reactivity via Convex. Using OpenAI’s Node SDK server-side, and Convex's useQuery hook client-side.
Moderating ChatGPT Content: Full-Stack
In this post, we’ll look at how to use the moderation API to flag messages before sending them to Chat-GPT, and patterns for handling these errors in a full-stack React app.
Adding Personality to ChatGPT-3
How to store multiple personalities Convex and provide them to the chatGPT API, enabling changing personalities mid-conversation. This is a follow-up to Building a full-stack ChatGPT app.
Use Convex to fetch an image from OpenAI’s image generation service based on a user-provided prompt.