New AI software development: Take off fast. Land safely. Learn more

Security Check
Start for Free Book a Demo Contact
English|German
Platform Login
Claude Code

Using Claude Code in your company.

Claude Code makes developers much faster. To turn that into good software, you need method and guardrails. We introduce your team to it and build alongside, in your repository and with your tools.

  • Onboarding and coaching
  • Security guardrails
  • Tests and reviews
Book an introduction call About Guided Flight
Definition

What is Claude Code?

Claude Code is an agentic coding tool by Anthropic. It reads the code of a project, edits files, runs commands and tests and checks the result, in dialogue with the developer, who can step in at any time.

It works in a loop and connects to other systems through interfaces such as MCP, for example ticket systems or databases.

Gather contextTake actionVerify the result

Where it runs

Terminal VS Code JetBrains IDEs Desktop app Browser

All versions share the same foundation. Project rules and settings apply everywhere.

Source: Claude Code documentation (retrieved in November 2026).

Use cases

What Claude Code is good for

We use Claude Code every day, for customer projects, our own AI platform and our internal AI Cockpits.

Building new features

From the data model to the interface, including tests. Ideal for AI Cockpits and applications that need to be in daily use quickly.

Understanding existing code

Have grown applications explained, modernize them and rebuild them step by step, without losing knowledge.

Ensuring quality

Write tests, narrow down bugs, check changes for vulnerabilities and document them.

Comparison

Claude Code or Cursor?

Both make developers faster with AI, but they take different approaches. We work with both and follow your team.

Claude Code

What it is
AI agent that works in the terminal, in the IDE, as a desktop app or in the browser
Approach
Agent first: describe the task, the agent plans, changes, tests
Models
Claude models by Anthropic
Billing
Claude subscription or usage via the API
Fits when
Teams want to hand larger tasks to an agent and automate them

Cursor

What it is
Code editor of its own with built-in AI and agent
Approach
Editor first: suggestions while typing, chat and agent in the editor
Models
Models from several providers to choose from
Billing
Cursor subscription with usage allowance
Fits when
Teams work closely in the editor and want to add AI step by step

Checked on 5 November 2026. Details: Claude Code, Cursor.

Costs

How much does Claude Code cost?

Claude Code is included in all paid Claude plans and shares the usage limits with the rest of the plan. Alternatively, you pay per use via the API.

PlanPrice (USD)For whom
Pro20 per month, 17 per month billed annuallyGetting started, occasional use
Max100 per monthDaily use, about 5 times as much as Pro
Ultra200 per monthIntensive use, successor to Max 20x
Team and Enterpriseper seat, see pricing pageCompanies with central administration
APIpay as you goAutomation, variable usage

Prices checked on 5 November 2026, without guarantee. Current prices: claude.com/pricing. Tip: if an API key is set on a computer, Claude Code uses it instead of the subscription and bills via the API (Anthropic Support).

Guardrails

What companies should pay attention to

Claude Code is only as good as the guardrails it works within. We clarify these points with every team first.

PermissionsWhich commands and files the agent may use without asking, and which never.
SecretsCredentials do not belong in the repository or in prompts.
Separate environmentsThe agent works with test data, not with production.
Review and testsEvery change is tested and reviewed before it goes live, including automatic checks for vulnerabilities.
Data protectionClarify which code and data go to the provider, and choose the right plan.
Project rulesWrite down architecture, conventions and security rules so that the agent follows them.
How we help

Faster with Claude Code, without compromising on security

We come from traditional software development and have been developing with AI since 2022. Whether your team builds itself, needs a foundation or wants reinforcement: choose your level.

Level 1 · You build

Guided Flight

Your team steers, we sit alongside.

Your teamEvoya AI
  • Training on your own project
  • Reviews of code and architecture
  • Guardrails for security and data
More about Guided Flight
Level 2 · Together

Ready for Takeoff

We lay the foundation, your team flies on.

Your teamEvoya AI
  • Database, login and backups
  • Your first AI Cockpit
  • Handover to your team
More about Ready for Takeoff
Level 3 · We buildFastest to the goal

AI Crew

We build with you, on site and at full speed.

Your teamEvoya AI
  • Developers with AI in your office
  • First version after a few days
  • Security review and handover
More about the AI Crew
Already built it yourself?Security CheckWe check your app for security, backups and architecture. View
With the tools your team uses:Claude CodeLovableCursorSupabase

Frequently asked questions about Claude Code

Claude Code is an agentic coding tool by Anthropic. It reads the code of a project, edits files, runs commands and tests and checks the result. It runs in the terminal, in VS Code and JetBrains IDEs, as a desktop app and in the browser.
Claude Code is included in all paid Claude plans: Pro from USD 20 per month, Max USD 100, Ultra USD 200 (checked on 5 November 2026). For companies there are Team and Enterprise plans, or pay-as-you-go billing via the API. Current prices: claude.com/pricing.
It depends on the team. Claude Code is agentic by design and suits handing over larger tasks. Cursor is a code editor with built-in AI and suits teams that work closely in the editor. Many teams use both.
Not necessarily for first results. For software that runs securely and maintainably in production, yes: someone has to be able to assess architecture, access rights, tests and backups. That is exactly what Guided Flight is for.
Yes, with clear guardrails: permissions for the agent, no secrets in the repository, separate test environments, review and tests before every release and a plan that matches your data protection requirements.
Yes, as a Guided Flight: we introduce your team to Claude Code using your own projects instead of exercises, set up project rules and security guardrails and review what gets built.

Should your team get faster with Claude Code?

In the introduction call we clarify where your team stands and which level fits: Guided Flight, Ready for Takeoff or AI Crew.

Book an introduction callRequest a Security Check

Claude Code is a product of Anthropic. Evoya AI is not affiliated with Anthropic. More on the topic: Vibe coding in companies · Lovable · AI software development