# What Agents Pick

> what AI coding agents install when the prompt names no tool. 355 runs of Claude Code on Opus 5.5, Sonnet 5.5 and Haiku 5.5, across 10 categories (database, ORM, auth, payments, email, deployment, background jobs, caching, file storage, error monitoring). every run starts from the same blank starter repo (Next.js + TypeScript or Python + FastAPI), the prompt asks for an outcome and never names a product, and the pick is read from the diff. data updated 2026-10-18. free to cite with a link, data is CC BY 4.0.

headline numbers use Sonnet 5.5. "built its own" means the agent added no third party package or service for the job and wrote it itself.

## results by category

- [Which database do coding agents pick?](https://whatagentspick.com/which-database-do-coding-agents-pick/): pooled SQLite file, 20 of 36 runs. Opus 5.5 Self-hosted Postgres, 8 of 12. Sonnet 5.5 SQLite file, 9 of 12. Haiku 5.5 SQLite file, 7 of 12. built its own 20 of 36.
- [Which ORM do coding agents pick?](https://whatagentspick.com/which-orm-do-coding-agents-pick/): pooled SQLAlchemy, 18 of 36 runs. Opus 5.5 SQLAlchemy, 6 of 12. Sonnet 5.5 SQLAlchemy, 6 of 12. Haiku 5.5 SQLAlchemy, 6 of 12. built its own 5 of 36.
- [Which auth provider do coding agents pick?](https://whatagentspick.com/which-auth-provider-do-coding-agents-pick/): pooled built its own, 35 of 35 (32 as hand-rolled JWT or sessions) runs. Opus 5.5 Hand-rolled JWT or sessions, 12 of 12. Sonnet 5.5 Hand-rolled JWT or sessions, 12 of 12. Haiku 5.5 built its own, 11 of 11 (8 as hand-rolled JWT or sessions). built its own 35 of 35.
- [Which deployment platform do coding agents pick?](https://whatagentspick.com/which-deployment-platform-do-coding-agents-pick/): pooled Plain Docker, 21 of 36 runs. Opus 5.5 a tie, Fly.io, Heroku and Plain Docker at 4 of 12 each. Sonnet 5.5 Plain Docker, 7 of 12. Haiku 5.5 Plain Docker, 10 of 12. built its own 21 of 36.
- [Which email API do coding agents pick?](https://whatagentspick.com/which-email-api-do-coding-agents-pick/): pooled Raw SMTP, 24 of 36 runs. Opus 5.5 Raw SMTP, 7 of 12. Sonnet 5.5 Raw SMTP, 10 of 12. Haiku 5.5 Raw SMTP, 7 of 12. built its own 24 of 36.
- [Which payments provider do coding agents pick?](https://whatagentspick.com/which-payments-provider-do-coding-agents-pick/): pooled Stripe, 35 of 35 runs. Opus 5.5 Stripe, 12 of 12. Sonnet 5.5 Stripe, 12 of 12. Haiku 5.5 Stripe, 11 of 11. built its own 0 of 35.
- [Which background job queue do coding agents pick?](https://whatagentspick.com/which-background-job-queue-do-coding-agents-pick/): pooled built its own, 33 of 33 (5 as hand-rolled queue or cron) runs. Opus 5.5 built its own, 12 of 12 (3 as hand-rolled queue or cron). Sonnet 5.5 built its own, 10 of 10 (2 as hand-rolled queue or cron). Haiku 5.5 built its own, 11 of 11. built its own 33 of 33.
- [Which cache do coding agents pick?](https://whatagentspick.com/which-cache-do-coding-agents-pick/): pooled built its own, 36 of 36 runs. Opus 5.5 built its own, 12 of 12. Sonnet 5.5 built its own, 12 of 12. Haiku 5.5 built its own, 12 of 12. built its own 36 of 36.
- [Which file storage do coding agents pick?](https://whatagentspick.com/which-file-storage-do-coding-agents-pick/): pooled Local disk, 34 of 36 runs. Opus 5.5 Local disk, 12 of 12. Sonnet 5.5 Local disk, 10 of 12. Haiku 5.5 Local disk, 12 of 12. built its own 34 of 36.
- [Which error monitoring tool do coding agents pick?](https://whatagentspick.com/which-error-monitoring-tool-do-coding-agents-pick/): pooled Sentry, 21 of 36 runs. Opus 5.5 built its own, 6 of 12. Sonnet 5.5 Sentry, 12 of 12. Haiku 5.5 Sentry, 5 of 12. built its own 11 of 36.

## questions

- [what stack does Claude Code pick by default?](https://whatagentspick.com/questions/what-stack-does-claude-code-pick-by-default/): on Sonnet 5.5, with no tool named in the prompt: database: SQLite file, 9 of 12. ORM: SQLAlchemy, 6 of 12. auth: Hand-rolled JWT or sessions, 12 of 12. deployment: Plain Docker, 7 of 12. email: Raw SMTP, 10 of 12. payments: Stripe, 12 of 12. jobs: built its own, 10 of 10 (2 as hand-rolled queue or cron). caching: built its own, 12 of 12. storage: Local disk, 10 of 12. monitoring: Sentry, 12 of 12. Stripe for payments is the only vendor every model picked every time.
- [do AI coding agents build their own instead of using a tool?](https://whatagentspick.com/questions/do-ai-coding-agents-build-their-own-instead-of-using-a-tool/): mostly yes. in 219 of 355 scored runs (62%) Claude Code added no third party tool for the job and wrote it itself. highest: auth 35 of 35, jobs 33 of 33, caching 36 of 36, storage 34 of 36. lowest: payments, 0 of 35.
- [which dev tools do AI coding agents actually install?](https://whatagentspick.com/questions/which-dev-tools-do-ai-coding-agents-install/): 12 of the 80 vendors we check for got installed at least once: Stripe 35 of 35 payments runs, Sentry 21 of 36 monitoring runs, SQLAlchemy 18 of 36 ORM runs, Self-hosted Postgres 16 of 36 database runs, Drizzle ORM 12 of 36 ORM runs, Resend 11 of 36 email runs, Fly.io 8 of 36 deployment runs, Heroku 7 of 36 deployment runs, PostHog 4 of 36 monitoring runs, Amazon S3 2 of 36 storage runs, Kysely 1 of 36 ORM runs, Postmark 1 of 36 email runs.
- [which dev tools do AI coding agents never pick?](https://whatagentspick.com/questions/which-dev-tools-do-ai-coding-agents-never-pick/): 68 of the 80 vendors we check for were never installed in 355 scored runs. that includes Clerk, Auth0, Supabase, Vercel, Prisma, Neon, Redis and SendGrid. some got named in the final message and still not installed, like Redis (12), Render (11), Vercel (9), SendGrid (9).

## model vs model

- [Opus 5.5 vs Sonnet 5.5](https://whatagentspick.com/compare/opus-5-5-vs-sonnet-5-5/): Opus 5.5 and Sonnet 5.5 lead with the same thing in 7 of 10 categories. they split on database, deployment, monitoring. database: Opus 5.5 Self-hosted Postgres, 8 of 12, Sonnet 5.5 SQLite file, 9 of 12. deployment: Opus 5.5 a tie, Fly.io, Heroku and Plain Docker at 4 of 12 each, Sonnet 5.5 Plain Docker, 7 of 12. monitoring: Opus 5.5 built its own, 6 of 12, Sonnet 5.5 Sentry, 12 of 12.
- [Opus 5.5 vs Haiku 5.5](https://whatagentspick.com/compare/opus-5-5-vs-haiku-5-5/): Opus 5.5 and Haiku 5.5 lead with the same thing in 7 of 10 categories. they split on database, deployment, monitoring. database: Opus 5.5 Self-hosted Postgres, 8 of 12, Haiku 5.5 SQLite file, 7 of 12. deployment: Opus 5.5 a tie, Fly.io, Heroku and Plain Docker at 4 of 12 each, Haiku 5.5 Plain Docker, 10 of 12. monitoring: Opus 5.5 built its own, 6 of 12, Haiku 5.5 Sentry, 5 of 12.
- [Sonnet 5.5 vs Haiku 5.5](https://whatagentspick.com/compare/sonnet-5-5-vs-haiku-5-5/): Sonnet 5.5 and Haiku 5.5 lead with the same thing in 10 of 10 categories. same prompts, same repos, same defaults.

## tools

- [every tool we check for](https://whatagentspick.com/tools/): 90 tools, one page each with counts per model, stack and prompt

## method and raw data

- [changelog](https://whatagentspick.com/changelog/): every round of runs, what each one changed
- [methodology](https://whatagentspick.com/methodology/): starter repos, prompt wordings, stop rules, how a pick is detected
- [manifest.json](https://whatagentspick.com/data/manifest.json): summary of every category and model
- [runs.json](https://whatagentspick.com/data/runs.json): every run with the evidence for its pick
- [full text version](https://whatagentspick.com/llms-full.txt): every number on the site in one file

## optional

- [newsletter](https://whatagentspick.com/newsletter/): one email when the picks move
- [about](https://whatagentspick.com/about/)