40+ AI models · Claude · GPT · Gemini · Grok · DeepSeek
Every Top AI Model. One Coding Agent.
Chat with Claude, GPT, Gemini, DeepSeek and 40+ more models in one place. The agent writes, runs and debugs code in a sandbox, searches the web, and works on your own files in the desktop app.
20 free credits every month, no card needed · Try the Unlimited plan free for 7 days
80,905+ Brilliant Minds
142.4B+ Tokens Processed
Web Search · Agent Mode
40+ LLMs · Sub-Agents
CodingFleet · New chat
online
New · Free desktop app
CodingFleet on your computer
An AI coding agent that works on your real code, in the folder you pick, and asks before it changes anything.
- Reads your project, runs your tests, edits your files
- Every command and file change waits for your OK, with one-click Undo
- Your plan's unlimited models work in the app, like in the chat
Windows may say “Windows protected your PC†until the app is signed: click More info → Run anyway.
The app is not signed by Apple. The first time, open it from System Settings → Privacy & Security → Open Anyway.
1,553,687
Generations so far
80,905
Happy Programmers
90+
Languages & Frameworks Supported
How It Works
From curiosity to shipping — in four quick steps.
Explore
Try the tools with zero sign-up. Get a feel for what effortless coding looks like.
Sign Up
Create a free account for 20 credits every month and access to 40+ models.
Chat or Install
Chat on the web, or get the free desktop app and let the agent work on your own code.
Ship
Boost velocity with AI that writes, reviews, tests, and explains — so you can focus on building.
Flexible Pricing Plans
Start free. Plans from $7/month, or unlimited standard models from $25/month with a 7-day free trial. One-time credit packs also available.
View Pricing PlansSee It in Action
One question, and it does the whole job.
Describe It in Chat, Get Working Code
Ask in plain English and it works the whole problem: pulls the live data, writes and runs the code in a sandbox, and hands back the charts, the files and a straight answer.
Start a ChatAnd Many More…
Code review, optimization, tests, docs, diagrams and an online code runner, in 90+ languages. There's always something new.
FAQs
The Code Assistant fixes your coding issues, improves your code, explains it, answers questions about it, or adds the features you ask for.
The Code Generator writes code in 90+ programming languages and frameworks from plain-language instructions. Be as specific as you can. Turn on Agent Mode to let it run and test the code in a sandbox before it answers.
The Code Converter converts your code from one programming language or framework to another, such as Django to Laravel or Python to C++. If your language or framework is not in the list, type it in the field below the code box.
The Code Explainer helps you understand code you do not know. You can set how detailed the explanation is.
The Documentation Generator adds docstrings and comments to your code, or writes documentation such as a README. Paste your code, or import a whole project, and choose how detailed the result is.
The Code Enhancer suggests several ways to improve your code. Select the suggestions you like and it writes the improved code for you.
The Code Reviewer reviews your code like a senior developer: it finds bugs, security problems, performance issues and style problems, and tells you how to fix each one.
The Code Optimizer makes your code faster and leaner without changing what it does. It shows the time and space complexity before and after, and in Agent Mode it runs both versions to prove the speedup.
The Code Runner is an online code editor and compiler. Write code in one of 12 languages, run it in the cloud, and ask the AI to explain, fix or improve it.
The Diagram Generator turns your code into diagrams such as flowcharts and class diagrams. It writes diagram code and runs it in a sandbox to draw the result. To go the other way, use Diagram to Code.
The Unit Test Generator writes test cases for your code, edge cases included.
The "Continue in Chat" button takes the result of a tool into the chat, where you can ask follow-up questions, request changes, or keep working on the code with any model.
Code Execution runs AI-generated code in an isolated cloud sandbox. It can create reports, charts, PDFs, Word files and more. You can upload files to the sandbox and work with them through your prompts. It works in every tool and supports many programming languages.
Web Access lets the AI model search the web and read public web pages. It helps with tasks such as web scraping, or checking the latest library and API documentation.
Parallel Agents (beta) in the chat let the AI start several sub-agents that work on separate parts of a task at the same time, then combine their results.
Chat Memory lets the AI look at your earlier conversations before it answers. It is available in the chat only.
Yes. The free CodingFleet desktop app for Windows, macOS and Linux lets the coding agent work directly on the files and projects on your computer. It uses the same credits as the chat.
The Agent API lets you run the CodingFleet coding agent from your own code, scripts and CI. Every paid plan includes it; higher plans can run more agents at the same time.
Yes. You can use the tools 3 times without signing up. With a free account you get 20 credits every month, plus free models that use no credits at all.
A credit is the unit CodingFleet uses to measure usage. Each request spends credits; how many depends on the model you choose and on how much text the request reads and writes. The model picker shows the cost of each model, and this page lists every model.
Small, fast models often cost about 1 credit per request, and the most powerful models cost more. Free models cost nothing. Free accounts get 20 credits on the 1st of every month. The Pro plan adds 250 permanent credits every month. The Unlimited plan and higher plans include unlimited use of standard models plus weekly premium credits that reset every Monday. One-time credit packs add permanent credits that never expire. Learn more about credits.
The Pro plan costs $7 per month and gives you 250 permanent credits every month. If you prefer to pay once, one-time credit packs are available:
- 200 credits for $7
- 500 credits for $12
- 1,000 credits for $19
- 2,000 credits for $35
- 5,000 credits for $69
- 10,000 credits for $129
- 20,000 credits for $209
The Pro plan costs $7 per month. It includes 250 credits every month that roll over and never expire, even if you cancel; Premium models; up to 200K input tokens per request (about 800,000 characters); 2 active sandboxes; the Agent API; and file uploads. Learn more.
Yes. New subscribers can try the Unlimited plan free for 7 days. You add a card at checkout and pay only when the trial ends. Cancel before the 7 days end and you pay nothing. One trial per person.
The Unlimited plan costs $25 per month or $259 per year, with a 7-day free trial. It includes unlimited use of standard models, 600 premium credits every week, Premium models, up to 200K input tokens per request (about 800,000 characters), 2 sandboxes with snapshots, the Agent API, and bring-your-own API keys. Learn more.
The Elite plan costs $45 per month or $449 per year. It includes unlimited use of standard models, 1,200 premium credits every week, Elite models, 5 requests a day on Claude Fable 5.1 and GPT-6 Astra, up to 400K input tokens per request (about 1.6 million characters), 5 sandboxes with a shell and file manager, the Agent API, and bring-your-own API keys. Learn more.
The Ultimate plan costs $99 per month or $989 per year. It includes unlimited use of standard models, 3,000 premium credits every week, Ultimate models, Claude Fable 5.1 and GPT-6 Astra with no daily limit, up to 800K input tokens per request (about 3.2 million characters), unlimited sandboxes, the Agent API, and bring-your-own API keys. Learn more.
The Ultimate Max plan costs $189 per month or $1,795 per year. It includes everything in Ultimate, with 6,000 premium credits every week and up to 1M input tokens per request (about 4 million characters, on models with a 1M-token context window). Learn more.
Every Monday at 00:00 (midnight) UTC. If you need more before then, you can buy a credit pack in the "Prefer a One-Time Purchase?" section of the pricing page. Pack credits are permanent and are used only after your weekly credits run out.
Buy a one-time credit pack below the plan cards on the pricing page. Pack credits never expire, and there are no recurring payments.
Your credits never expire, and your account permanently gets the Pro features: Premium models, up to 200K input tokens per request (about 800,000 characters), and file uploads.
Yes. Stripe handles all payments. It is fully encrypted and complies with industry standards; CodingFleet never sees your full card number.
Yes, you can cancel at any time from your Billing page. If you need help, contact us.
Your credits never expire and stay in your account, even after you cancel.
You keep the plan features until the end of the period you paid for. Then your account moves to the Free plan. Credits from credit packs stay in your account.
We offer a 14-day money-back guarantee. If you are not satisfied within the first 14 days of your Unlimited, Elite, Ultimate or Ultimate Max subscription and have used the tools no more than 10 times, contact us and we will give you a full refund.