Technology

In very simple terms, a token is more or less a word. You pay per input and output tokens (your prompts and the answers) as they correlate the most closely to the energy expended by the LLM to process your request.

FYI the flash model is ~158 GB

If you use it for Q&A, that's a lot of tokens. If you use it to write software somewhat autonomously, it's easy to go through a million tokens every few hours. Do that every day and you'll be paying over $100 a month at that rate.

The ones paying attention and on a budget would still use them. "The best" of anything is usually not cost effective.

Even before reducing the prices, they were already 2 to 3 times cheaper than equivalent alternatives from Anthropic's ($3in, $15out) and OpenAI's ($1.75in, $14out) at $1.74in and $3.48out. Now they're around 10x cheaper.

Edit: Deepseek V4 Flash is the leading model on OpenRouter

so much for no one buying it

Also, the big models are in the infant stages of turning the monetization screws. This could be a tactic to knock their owners' legs out from under them. If they can't turn up their prices as much or as fast as they had hoped due to the competitive pressure, venture capital may begin to divest sooner. At their current debt levels, that would literally end OpenAI and I believe also Anthropic.

If a temporary "permanent" price cut wipes out some competitors, that will make for smoother sailing in the future for them to raise prices more than they otherwise would have been able to, quickly recouping the cost of the price reduction and then all profit after that.