Color glitches with marker_z and GR backend

[axsk]

Details

Using Plots.jl from the REPL through the connected VSCode plots pane, after a while my scatter plots colors start to behave strangely, with some points showing in glitchy colors which are not even contained in the palette.

I will attach a screenshot the next time this occurs.

Backends

I observed this bug on GR only, at which point I switch to plotly() which then works.
Switching back to gr() the problem remains.

Versions

Plots.jl version: v1.41.6
Backend version (]st -m <backend(s)>): GR v0.73.24
Output of versioninfo():

Julia Version 1.12.5
Commit 5fe89b8ddc1 (2026-02-09 16:05 UTC)
Build Info:
  Official https://julialang.org release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 128 × AMD EPYC 9334 32-Core Processor
  WORD_SIZE: 64
  LLVM: libLLVM-18.1.7 (ORCJIT, znver4)
  GC: Built with stock GC
Threads: 8 default, 1 interactive, 8 GC (on 128 virtual cores)
Environment:
  JULIA_CUDA_SOFT_MEMORY_LIMIT = 10G
  JULIA_HISTORY = ./.history.jl_

Edit:
Here we have some offending pink in the viridis scheme (the red marker is intentional)

And here some cyan and blue

[axsk]

A workaround is Plots.GR.reset()

[maetshju]

I am also encountering this issue. Here's a minimal script to reproduce:

using Pkg
Pkg.add("Plots")
using Random, Plots
Random.seed!(9)
x = rand(1000)
y = rand(1000)
z = rand(1000)
plot(x, y, marker_z=z, seriestype=:scatter, color=:jet)
savefig("scatterjet.png")
plot(x, y, marker_z=z, seriestype=:scatter, color=:inferno)
savefig("scatterinferno.png")
plot(x, y, marker_z=z, seriestype=:scatter, color=:binary)
savefig("scatterbinary.png")

Running this with a new temp environment via julia --project=@temp pulls down the newest Plots.jl, v1.41.6 and produces the following images, with the error apparent in the binary colorscheme. Note that this error does not occur if the second plot is not executed.

Jet result

Inferno result

Binary result (which should be all grayscale)