What is the best strategy for solving captcha with/without proxy when parsing Yandex search results?

Question

Search Engine Optimization

There are, let's say, 100,000 search phrases for which you need to parse search results from Yandex. Some program is taken that can do this (for example, Key Collector) and the parsing process begins, but Yandex slips its captcha and money in services will be spent on its recognition. So, a certain amount can accumulate for captcha, but how can it be reduced as much as possible? Maybe I don’t know something, or somehow I’m setting up the parsing process in the wrong way and for captcha it’s necessary to use proxy servers (which also cost money), or set a multi-second delay between requests so that Yandex does not slip the captcha?
Let's say parsing 100,000 requests from one account can take a week. And for several accounts, you will need to buy a proxy (otherwise Yandex will suspect something if the requests are from different accounts, but from the same ip, and again slip the captcha).
I admit that it is possible that there are even options for using free software to solve captcha, but so far I have only found xevil, in which the free version, as it turned out, does not solve Yandex captcha, and the paid one costs 14,000 rubles and this is no good for small volumes.

Answer 1

To prevent captcha from disturbing the key collector:
1. Select accounts and check them for captcha. Leave only those that will be without captcha.
2. Buy a proxy (you can find it for 100 rubles apiece https://proxy-sale.com/russian-proxy.html ). At least 5 pieces.
3. Set the time to 15 - 20 seconds.

Answer 2

For such a volume, you need to take several thousand resident proxies, try to guess captcha with software like xevil or kamonstr if they know how to guess Yandex captcha, fixed costs come out