4 ms·
It’s shocking how close this feels to claude, obviously it's much slower, but I don’t know that it’s significantly dumber. Interestingly the imatrix quantizatio
by FuckButtons 5mo ago
It’s shocking how close this feels to claude, obviously it's much slower, but I don’t know that it’s significantly dumber. Interestingly the imatrix quantization seems to be better than whatever quant the zdr inference backends on open router are using. It was self aware enough yesterday to realize that it’s own server process was itself without me telling it, which is not something I’ve ever observed a local model doing before.
- stavros 5mo agoIn my (obviously anecdotal) testing, DeepseekV4 Pro was better than Sonnet at coding. However, it is much slower, but also many times cheaper, especially with the promotion right now.
- DeathArrow 5mo agoDo they have a coding plan or you only pay per API call?
- trollbridge 5mo agoIt’s just per token, but burning up 100 million+ tokens is a $3 transaction with their pricing right now
- DeathArrow 5mo agoDo you use the official API or another provider?
- stavros 5mo agoI use the official API, OpenRouter somehow didn't use caching and one short session with Qwen cost me $5.
- trollbridge 5mo agoJust directly. Paid for it with PayPal. It’s quite simple to set up and use.
- ReptileMan 5mo agoYou pay per api call but you will be challenged to burn trough 20$ per month. 24/7 usage for single agent will probably cost you around 100$ per month. It is very efficient especially with modern harnesses.
- thejazzman 5mo agoI racked up $30 in 3 days, but I did A LOT of refactoring. Got my projects really buttoned up and now I’m sipping tokens with codex again. Have been more like $1-2/day with deepseek since that initial swarm. With max effort. It’s especially great that you don’t have to worry about hitting your limit and being stalled. I’m using it with Claude
- redman25 5mo agoWhat prompt had you given it?