Recently, many token-selling providers have been raising prices in disguised ways. Just as everyone was getting used to “AI is getting more expensive,” DeepSeek did something absurd — they permanently cut the price of V4-Pro to a quarter of the original.
In fact, the pricing when DeepSeek V4 launched was already impressive. V4-Flash was cheap to begin with, and with cache hits, the cached token cost was almost negligible. Pro was much more expensive, but at the launch 75% discount, it was on par with Minimax’s pricing. And Minimax has smaller parameters and only a quarter of the context window. So even at the original price, DeepSeek Pro still offered great value.
But now that DeepSeek has permanently kept the Pro price at that level, Minimax looks like a terrible deal. Besides the gap in parameters and context, Minimax’s Token Plan has been almost guaranteed to hit peak-hour throttling after 2 PM over the past month. Can’t use it when you need it — value means nothing.
DeepSeek, on the other hand, maintains low pricing, decent availability, and rock-solid stability. Hard not to switch. I imagine many individual users feel the same way.
I hope this state of high value and stability continues. After all, without AI, my work beyond literature reading would become a lot harder.
