Developer Tools, Ranked

Three pricing physics in one category

Developer tools bill in three different physics, and comparing them on sticker price mixes units. Per-seat tools (GitHub at $4, Postman at $14) scale with headcount: predictable, and expensive exactly when occasional users pile up. Usage-metered tools (Sentry on events, CircleCI on build minutes) scale with activity: cheap in quiet months, spiky when production misbehaves, which is precisely when you cannot turn them off. Platform tools (Vercel, GitLab) bundle several jobs into one subscription where the entry seat is honest but the growth axis (bandwidth, minutes, storage) is where the money is. The table normalises all three to a published entry price; the prose is the warning label.

The self-hosting axis actually matters here

More than in any other category, developer tooling can be owned outright. GitLab publishes a full self-managed edition and Sentry ships self-hosted; both grade at the top of the lock-in column for it. GitHub Enterprise Server exists but is an enterprise motion; Vercel and Postman are cloud-only. For most teams the managed cloud is the right default, but the option changes negotiating posture: a team that can credibly migrate to self-hosted GitLab renews on different terms than one that cannot. The grade encodes that leverage, not just migration pain.

Where the free tiers stop being free

Every tool here has a real free tier, because the category competes on developer adoption, and the on-ramp is the marketing. The ceilings are consistent: collaboration and control, not usage. GitHub's free private repos meet a wall at protected branches and required reviewers. Sentry's 5k events vanish on the first noisy release. Vercel's hobby tier excludes commercial use outright. CircleCI's free minutes cover a side project, not a monorepo. The pattern to plan around: the upgrade trigger will be a team-shaped feature, it will arrive suddenly, and it will land in the same sprint as whatever incident exposed it. Budget the paid tier from the day the second engineer joins, and the free tier becomes a genuine trial instead of a debt.

Where each of the six earns its slot

GitHub, at $4 the cheapest paid seat in the table, wins on gravity: the ecosystem, the contractors who already have accounts, and a rare pairing of high marks on both editorial axes. GitLab's $29 Premium seat buys the opposite thesis: repos, CI/CD and registry as one unit you can fully self-host, which is why it posts the category's highest power rating and its best lock-in posture at once. Sentry is the only monitoring entry and behaves like infrastructure: self-hostable, full export, priced by events rather than people. Vercel holds the simplicity end (the highest such rating among the six) precisely because it absorbs the deploy pipeline into itself, which is also why its export story is the thinnest. Postman, at $14 per seat, is the quiet workhorse: API development and testing with a power rating that surprises people who think of it as a request client. CircleCI is the specialist, selling build minutes as the product, and earns its slot in teams whose CI needs outgrow what their forge bundles.

Assembling a stack instead of picking a winner

Unlike most categories on this site, nobody picks one developer tool; the real question is which two or three compose into a stack without double-paying. The common seams: GitLab already contains CI, so pairing it with CircleCI means paying twice for minutes unless a specific runner need justifies it; GitHub plus Vercel is the lowest-friction pairing for web work, with Actions covering what Vercel does not absorb; Sentry composes with everything, since none of the forges does real error monitoring. A three-seat team can run GitHub, Vercel's hobby-to-Pro path and Sentry's entry tier for under $100 a month at published prices, and the table above is the receipt. The stack-level warning is concentration risk: every tool you add from one vendor improves the bundle price and worsens the exit, and the lock-in column is per-tool, not per-vendor. Read it with that in mind, and re-read it at renewal time: the discount that consolidation buys this year is priced against the leverage it removes from every negotiation after.