You would use materials that perform completely fine at those temps. This could be anything from high nickel alloy steel, to Inconel, to an HEA (high entropy alloy). You can even do high heat resistant metals with ceramic coatings on the inside for protection if creating a passivation layer is too difficult for the application or the exposure environment does not allow for one to form.

There is an entire subsection of engineering studies focused on purely coaxing specific properties out of a material or developing materials to custom suit extreme applications, known as material science. They generally work very closely with chemical engineers (my background) and metallurgists in order to manufacture the designed product in either batch form, or in continuous fashion.

I work in a steel mill and we have Inconel furnace rolls that hang out in 1600 F heat 24/7 and are rated (iirc) to ~2300F max operation temp. For reference medium carbon steel melts between 2600 and 2800F, and loses a lot of its mechanical strength well before 2300F (I am trying to find a stress strain curve for carbon steel over multiple temperatures for reference. I will update if I find one)

Edit: Okay so I found one that does show what I am trying to convey. As you can see, the higher the temperature of the sample material, the lower the yield strength. Example: the 100C sample was strained to >25% before failure, while the 700C sample began to plastically deform (fail) before 10% strain. Take note of the second link, all the test temperatures are MUCH higher than any of the carbon steel samples

Carbon Steel Curve: https://www.researchgate.net/figure/Stress-strain-curves-at-different-temperatures-for-steel-4509-2_fig11_236341600

Inconel Curve: https://www.researchgate.net/figure/Stress-strain-curves-of-Inconel-625-alloy_fig11_338984803