Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- I have been trying to stick with a DeepSeek+ OpenCode setup for the whole week, but seems like my subsidized Claude Max is still quite relevantby krlx
- Also, just talking about the raw input / output raw cost leaves out half of the equation. You also have to think about how a given providers caching strategy behaves, and how many tokens a given model actually uses to complete a task on average.
Just because a model has a higher input / output token cost doesn't necessarily always mean it's going to be more expensive to use.
by elliottshort - I don’t think you could’ve picked a shittier title.
Don’t bother with the link, here’s a quick rundown:
DeepSeek is raising API prices for its V4 models and introducing peak/off-peak pricing. For V4-Pro, uncached input goes from $0.435 per million tokens to $0.66 off-peak or $1.32 at peak; output rises from $0.87 to $1.98 or $3.96. The four-digit percentage increase comes from cached input, which was priced at an unusually low $0.003625 per million tokens and is rising to as much as $0.044.
Even after the increase, DeepSeek remains cheap relative to major frontier models. Claude Sonnet 5 costs $2 per million input tokens and $10 per million output tokens; Claude Opus 5 costs $5/$25. OpenAI’s GPT-5.6 Terra is $2/$12, while GPT-5.6 Sol is $5/$30.
So at peak rates, DeepSeek V4-Pro still costs about two-thirds as much as Sonnet on input and roughly 40% as much on output. Off-peak, the gap is wider.
by cloudie78