Pricing

Starter

For one developer, with flexible token usage and extra tokens billed per token.

  • Diff-only review on every pull request
  • Findings that name the function and quote the line
  • A probability on every finding
  • Billed on input tokens only

€5 / month

Choose Starter

Enterprise

For a company with many repositories and its own rules on retention.

  • Everything in Pro
  • Per-scan billing, no seat licences
  • A review of your retention rules

Pay as you go

Contact us

Cost per bug found, against the scanners we measured

SystemPrice / bugBugs found
DeepSeek V4 Flash$0.000524 of 32
Jevlit$0.000715 of 32
GPT-5.6 Luna$0.000716 of 32
MiniMax M2.7$0.001412 of 32
Kimi K2.5$0.001617 of 32
GLM 5$0.002717 of 32
Kimi K2.6$0.004020 of 32
Kimi K2.7$0.004522 of 32
DeepSeek V4 Pro$0.005922 of 32
Kimi K3$0.007323 of 32
GLM 5.2$0.007316 of 32
GPT-5.6 Terra$0.007819 of 32
GLM 5.1$0.009821 of 32
Gemini 3.5 Flash$0.01814 of 32
GLM 5.3$0.02026 of 32
Claude Sonnet 5$0.02519 of 32
Claude Opus 5$0.03728 of 32
GPT-5.5$0.06024 of 32
GPT-Daybreak Blue$0.08430 of 32
GPT-5.6 Sol$0.08623 of 32
GPT-6 Astra$0.13024 of 32

One scan of dsvpwa per system, priced per labelled vulnerability it actually found. Public rows are scored with the CWE check; Jevlit does not output CWEs, so its row is matched by file and line only. We are not top of this table and it is not sorted to make us look like we are: dsvpwa is a deliberately vulnerable application, where flagging every function scores 20 of 32 on its own. Measured by us, one run each; your repository will differ.

Every plan includes

  • No config file and no workflow of its own to maintain
  • Removing Jevlit later leaves nothing behind
  • The Claude Code plugin, for a verdict before you commit
  • No seat licences and no minimum spend