204 AI coding assistants in one list: the established products, and every new one as it launches.
Updated
82 apps launched in the last 12 months, by month.
The latest to launch, newest first.
| # | App | Launched |
|---|---|---|
| 1 | AU Autohand Governed autonomous engineering turning human intent into software | * |
| 2 |  CoPatch Unify your team and AI agents in one secure, shared workspace | * |
| 3 |  WormGPT Chat with an uncensored AI that keeps no logs and requires no KYC | * |
| 4 |  ArcBrain Add persistent local memory to your AI coding assistant | * |
| 5 |  Orgn ORGN CDE is the enterprise AI IDE for developers who require privacy. A powerful AI code generator and software development tool guaranteeing security with confidential computing, provable encryption, and zero data retention. | * |
| 6 |  CSSPicker Copy CSS from any site and generate clean UI code with AI | * |
| 7 |  Claudify Supercharge Claude Code with agents, 1,700+ skills, and persistent memory | * |
| 8 | PI Pinacle Start instant browser-based dev environments for serious vibe-coding | * |
* Launch date estimated.
The products every new one is compared with.
Grouped by what each app says it does. 79 of 204 apps fit a group so far.
| Type | Apps | Share | Examples |
|---|---|---|---|
| AI code assistants | 31 | 39.2% | A9X11 IDE Assistant, AI Code Generator & Assistant, AI Code Generator: Code Editor |
| Agent orchestration | 18 | 22.8% | Acyclic, AgSoft, ArcBrain |
| Autonomous coding agents | 14 | 17.7% | Agen, AgenKit, Anycode AI |
| Code review and collaboration | 6 | 7.6% | Capy, Endstack, Luke |
| Domain-specific assistants | 6 | 7.6% | Android AI, Banana Code, Datrics |
| Agent models and infrastructure | 3 | 3.8% | Claude Capybara, GLM-4.5 Open-Source Agentic AI Model, MEGA Code |
| Code security | 1 | 1.3% | Epherem |
What we could detect on their websites.
| Framework | Apps | Share |
|---|---|---|
| Next.js | 78 | 81.3% |
| SvelteKit | 4 | 4.2% |
| Astro | 3 | 3.1% |
| PHP | 3 | 3.1% |
| Remix | 3 | 3.1% |
| WordPress | 3 | 3.1% |
Of 96 apps whose framework we could detect.
| Host | Apps | Share |
|---|---|---|
| Vercel | 79 | 95.2% |
| Netlify | 4 | 4.8% |
Of 83 apps whose host we could detect.
How many of the 204 apps show a price, and how many sell to large companies.
| Measure | Apps | Share of 204 |
|---|---|---|
| Publish a pricing page | 62 | 30.4% |
| Offer an enterprise plan | 6 | 2.9% |
“ai coding assistant tool” is searched 110 times a month on Google in the US. These are the searches around it.
| Search | Per month | Cost per click |
|---|---|---|
| gemini code assist | 6,600 | $16.15 |
| codeium | 4,400 | $19.95 |
| ai code generator | 3,600 | $13.39 |
| best ai for coding free | 590 | $10.00 |
| free ai code generator | 590 | $10.91 |
| free ai coding assistant | 260 | $9.69 |
| best ai for coding free unlimited | 90 | $13.22 |
| free ai code generator without login | 90 | $5.97 |
| free unlimited ai code generator | 70 | $9.31 |
| ai code generator python | 50 | $10.18 |
| free ai code generator online | 40 | $9.89 |
| free ai coding agent for vs code | 30 | $12.72 |
| best free ai for coding 2026 | 20 | $9.56 |
| ai code generator app | 10 | $9.62 |
| ai code generator claude | 10 | $5.87 |
Cost per click is what advertisers bid for one visit from that search. Figures from Oct 2026.
An AI coding assistant helps write, understand, and change code from natural language prompts and the context of your project. It runs inside an editor, a terminal, or a browser pane, and it watches what you type to suggest code that fits your style and stack.
The assistant builds context before it writes anything. It indexes key files, open tabs, error messages, and tests, and may keep a lightweight map of symbols and dependencies in your repo. Tools like Codeium integrate with editors to surface this context, while Aider focuses on mapping and navigating a codebase from the command line.
A large language model turns your goal into code. It predicts likely code tokens and explanations, conditioned on your files and instructions. Providers like Anthropic supply the underlying models that power chat, completion, and reasoning.
Some assistants act like an agent. With your approval they can create or edit files, run commands, and iterate until checks pass. Cursor and Cline follow this agent pattern, proposing diffs and command plans you can accept or modify.
Results appear as inline completions, chat answers with code blocks, or patch-style diffs. For greenfield work, they can scaffold a project or component from a short brief; Shipixen focuses on producing ready-to-run codebases and handling deployment steps.
There is a feedback loop. You review, run tests, and paste failures back. The assistant explains stack traces, proposes fixes, and repeats until it compiles and the suite passes. Over time it adapts to your conventions in Python or JavaScript, so suggestions feel idiomatic.
Vendors combine access to models, editor integrations, and governance features into simple plans.
| Pricing model | How it works |
|---|---|
| Per seat | Each person with an account adds to the bill. |
| Usage based | Charges scale with generated tokens, requests, or build minutes. |
| Tiered subscription | Feature bundles for individuals, teams, and enterprises with higher context limits and admin controls. |
| Per workspace or org | A flat rate for a project, repository, or organization regardless of individual users. |
| Self-hosted license | A fixed subscription to run the server or models in your own environment. |
| Platform plus BYO model | A platform fee while you pay a separate model provider like Anthropic directly. |
