This is a complex problem with a lot of parts to consider, but I will attempt to address it with a back of the envelope evaluation. Global CO2 emissions were approximately 2.5 billion metric tons in 1900 and as of 2020 were around 33 billion metric tons according to the Center for Climate and Energy Solutions (https://www.weforum.org/agenda/2022/11/visualizing-changes-carbon-dioxide-emissions-since-1900/). CO2 emissions account for 76% of the total greenhouse gas emissions worldwide. In order to reach 1900 levels we would need to sequester around 30.5 billion metric tons per year.

It is estimated that about 55% of our current level of emissions are absorbed back or sequestered naturally through plant respiration and absorption into our oceans (which causes its own problems), but that leaves us with about 45% of 30.5 billion tons to sequester manually or about 13.725 billion tons. There was a recent technique that came out which is the cheapest method of sequestering carbon as of right now at $39 USD/ metric ton (https://www.cnbc.com/2023/01/24/new-technique-from-us-national-lab-to-remove-co2-at-record-low-cost.html). To capture that carbon, the top of the line models will use up to 30% of a power plants capacity to reduce 90% of the carbon output. Given that the majority of direct CO2 emissions are from power plants, lets simplify the problem a little bit by assuming that all carbon emissions come from power plants and to reduce carbon emissions we use this carbon capture technology.

EPA estimates that approximately 884.2 lbs of CO2 are emitted per megawatt-hour for a coal fired power plant (https://www.epa.gov/energy/greenhouse-gases-equivalencies-calculator-calculations-and-references#:~:text=The%20national%20average%20carbon%20dioxide,EPA%202021%3B%20EIA%202020b). For the 13.725 billion tons we discussed before, that would be about 15,522,506 mega-watt hours of generation capacity. Assuming the 30% from before, it would take 4,656,751 mega-watt hours of power to reduce the 13.725 billion tons by 90% (which would bring us down to 1.23525 billion tons). With the cost estimate of $39 USD/metric ton from before, it would cost about $482 billion USD and that is just operating costs, not including upfront costs of manufacturing the amount of systems that would be needed and installing them.

I need to throw in a disclaimer that this is a gross simplification of the problem and please don't quote me on this because I had to make a lot of simplifying assumptions.