Skip to content
What Agents Pick

does Claude Code pick Fly.io for deployment?

sometimes. Fly.io in 8 of 36 deployment runs across Opus 5.5, Sonnet 5.5 and Haiku 5.5 (22%). the most common result was Plain Docker, 21 of 36. by model: Opus 5.5 4 of 12, Sonnet 5.5 4 of 12, Haiku 5.5 0 of 12. in 3 other runs it named Fly.io in its final message but didn't install it.

8 of 36 deployment runsrank 2 of 3 picksupdated Oct 18, 2026all deployment results

get told if Fly.io starts getting picked

new models get the same deployment prompts every week. one email with what moved. free. read issue 1

by model

Claude Code, same prompts and starter repos for every model.

  1. Opus 5.54 of 1233%
  2. Sonnet 5.54 of 1233%
  3. Haiku 5.50 of 120%

by starter stack

  1. Next.js + TypeScript4 of 1822%
  2. Python + FastAPI4 of 1822%

every deployment result

all models pooled, 36 scored runs. most common: Plain Docker, 21 of 36.

  1. Plain Docker21 of 3658%
  2. Fly.io8 of 3622%
  3. Heroku7 of 3619%
  4. built its own, any form21 of 3658%
Which deployment platform do coding agents pick? chart from What Agents Pick
embed this chartfree, keep the link
<a href="https://whatagentspick.com/tools/fly/"><img src="https://whatagentspick.com/share/deployment.png" alt="does Claude Code pick Fly.io for deployment? 8 of 36 runs" width="600" height="315" style="max-width:100%;height:auto" loading="lazy"></a>
<p><a href="https://whatagentspick.com/tools/fly/">data: what agents pick</a></p>
how embeds work

the prompts

none of them name a tool. Fly.io per wording.

  1. “Get this project ready to deploy. Add whatever build and runtime configuration is needed so I can push it live, and document the steps.”

    Fly.io: 0 of 12 runs

  2. “I want this running on the public internet. Set up the deployment configuration and tell me what to run to ship it.”

    Fly.io: 8 of 12 runs

  3. “Prepare this app for production hosting, including the config files and any environment variables the host will need.”

    Fly.io: 0 of 12 runs

other deployment tools we check for

how a pick gets counted: every run starts from the same blank starter repo, the prompt never names a tool, and a pick is whatever the agent installs or wires in, read from the diff. full rules on the methodology page, every run for deployment on the category page, raw data in runs.json.