← Back to Blog

Dark or Light? A Codex Theme Color Preference Survey

2026-10-25

Codex theme color preference is one of those questions everyone has an answer to and almost nobody has data on. We ran a two week poll in the community channel and the theme gallery, collected 412 responses, and the headline number turned out to be the least interesting thing in the set.

Here is the headline anyway: 71 percent run dark as their default. The useful part starts after that number, because once you break it down the split stops looking like a preference and starts looking like a set of working conditions.

What the survey actually asked

Three questions, deliberately boring. Which mode you run by default, when you last switched, and what made you switch. No rating scales, and no ranking themes by how good they look.

  • Default mode: dark, light, or matched to the system setting
  • Last switch: this week, this month, this quarter, longer ago
  • Trigger: room lighting, screen sharing, eye strain, new display, season
  • How many themes are in rotation right now

The numbers behind codex theme color preference

The raw split was 293 dark, 87 light, 32 matched to the system. Break it down by how recently people switched and the picture changes: the dark group is large and mostly static, while the light group is small and moves around constantly.

  • Dark as default: 293 of 412 responses, or 71 percent
  • Light as default: 87, or 21 percent
  • Matched to the system setting: 32, or 8 percent
  • Dark users who switched modes in the last month: 24 percent
  • Light users who switched modes in the last month: 61 percent
  • Median number of themes in rotation: 3

Read that way, developers preferring dark is really a statement about where people settle. It says very little about what they enjoy, and almost all of the movement sits in the light group.

What actually makes people switch

The trigger question was the one worth asking. Almost nobody switches because they got bored of a look. They switch because something in the room or on the desk changed.

  • Room lighting: 38 percent, the largest single trigger
  • Sharing the screen in a meeting or a demo: 26 percent
  • Eye strain after a long stretch of work: 19 percent
  • A new monitor with different brightness or panel type: 11 percent
  • Genuinely changed their mind about the look: 6 percent

Why the headline number is not much use to you

If you are choosing one theme for yourself, knowing that 71 percent of people run dark tells you nothing about your room, your monitor or the hours you keep. It is a fact about a group you are not required to join.

It does explain one thing: why the gallery is mostly dark. If you run light mode and you are looking for company, you are in a real minority, and you should expect to do more of the tuning yourself.

How to run this on your own setup

Cheaper than a poll and far more accurate for you. Two weeks, one line a day, and you end up with better data about your own eyes than any survey can hand you.

  • Pick a mode and stay in it for a week without switching
  • When you feel like switching, note the time of day and the room lighting
  • Note whether you were sharing your screen at that moment
  • At the end of two weeks, keep whichever mode you were least tempted to leave

Whichever way it lands, switching cleanly is covered at /en/blog/best-practices-for-switching-codex-themes-without-breaking-flow, and the two modes side by side are described at /en/blog/codex-theme-light-dark-modes.

Do most developers prefer dark Codex themes?

In this survey, 71 percent ran dark as their default. The more telling number is that only 24 percent of them had switched in the last month, so dark is mostly where people settle rather than what they actively pick.

What makes developers switch between dark and light?

Room lighting was the largest trigger at 38 percent, followed by sharing the screen at 26 percent and eye strain at 19 percent. Very few people switch because they changed their mind about the look.

Should I use a dark theme because most people do?

No. Your monitor brightness, your room lighting and the hours you work matter more than a population average. Run the two week test instead and keep whichever mode you were least tempted to leave.

How many themes do people keep installed?

Three was the median. Most respondents had one default, one for screen sharing, and one they were still evaluating.

Related Posts