Tool category
Coding & development
Autocomplete, chat over your codebase and autonomous coding agents.
AI coding assistants complete code, explain unfamiliar projects and handle whole development tasks largely on their own. For learners and hobby developers there are generous free offers; anyone who codes daily benefits from the more powerful paid versions. Even without deep prior knowledge they help you build small projects.
How do I choose the right tool?
- For learning and small projects, the free offers are enough.
- Those who develop daily benefit from an agent for larger tasks.
- Check the integration into your editor (e.g. VS Code).
- For sensitive code, review the privacy and data-usage rules.
Best free tools
Permanently usable free tier – try it at no cost.
- 1
GitHub Copilot FreeGitHub/Microsoft
Code completion & chat in VS Code, JetBrains etc. – limited quota
Boilerplate, Explaining code, Writing tests
Free (2k completions/mo)Free tierOpen - 2
OpenAI Codex (Free)OpenAI
OpenAI’s coding agent in the browser, terminal and IDE – also in ChatGPT’s free plan with a limited quota
Building features, Bug fixes, Explaining code
Free (5-hour window)Free tierOpen - 3
Google AntigravityGoogle
Google’s free agent IDE and CLI (successor to Gemini Code Assist) with current Gemini models
Everyday coding, Reviews, Refactoring
Free (weekly agent limit)Free tierOpen - 4
Devin Desktop (Free)Cognition
AI IDE by Cognition (formerly Windsurf) with a local agent, understands whole projects
Project-wide changes, Bug fixes, Code questions in context
Free (unlimited Tab)Free tierOpen - 5
Cursor HobbyAnysphere
Popular AI IDE to try out, with limited requests
Testing the IDE, Small projects, Code explanations
Free (limited)Free tierOpen - 6
Amazon Q DeveloperAWS
Completion, chat and security scans, strong in the AWS ecosystem
AWS projects, Code scans, Cloud architecture questions
Free (50 agent requests/mo)Free tierOpen
Best premium tools
Monthly prices as entry and maximum plan (approximate).
- 1
Claude (Pro/Max) + Claude CodeAnthropic
Terminal agent for entire tasks: plans, edits multiple files, runs tests
Building features, Refactorings, Understanding legacy code, Code reviews
From approx. 20 €Max approx. 200 €/moOpen - 2
Cursor Pro/UltraAnysphere
AI IDE with multi-file edits, agent mode and access to top models
Daily coding, Large codebases, Fast prototypes
From 20 $Max 200 $/moOpen - 3
Codex (ChatGPT Plus/Pro)OpenAI
Full Codex quota in web, CLI and IDE, parallel cloud tasks; Pro tiers for heavy users
Agentic development, Parallel tasks, Code reviews
From 20 $Max 500 $/moOpen - 4
GitHub Copilot Pro/Pro+GitHub/Microsoft
Unlimited completion, coding agent for issues/PRs, model choice
PR automation, Issue → code, Team standard
From 10 $Max 100 $/moOpen - 5
Devin Pro / MaxCognition
Devin Desktop plus cloud agents: full agent quota, all models, parallel sessions
Agentic work, Team rollout, Large refactorings
From 20 $Max 200 $/moOpen - 6
Replit CoreReplit
Replit agent builds & deploys apps entirely in the browser incl. database
Apps without setup, Learning projects, Quick deployments
From 20 $Max 100 $/moOpen
Matching guides & automations
Build a working app without coding skills
From idea through UI design and code to a published web app with database.
Test a product idea up to a prototype
From the market check through the concept and product images to a test landing page.
Loop engineering: tasks that run themselves Advanced
Instead of prompting the AI step by step, you give it a measurable goal – and it works round after round, checks itself and keeps going until the goal is met. Made for recurring or multi-part work. Note: more demanding than the other cases – requires Claude Code (command line) and a paid Claude plan.