Re:DeepSeek V4–almost on the frontier, a fraction of the price
by try-working on Friday May 01, 2026 @10:17PM
(#47982702)
Someone on Twitter got >200M tokens for around $10 at the current pricing level
Re:DeepSeek V4–almost on the frontier, a fraction of the price
by rvz on Saturday May 02, 2026 @12:51AM
(#47983431)
So it begins.
Re:DeepSeek V4–almost on the frontier, a fraction of the price
by aitchnyu on Saturday May 02, 2026 @04:50AM
(#47984655)
What would be the non-subsidized price for a V4 api? Can it be priced 3x cheaper than bigger models? In Openrouter, this 1600B param model costs 0.4$. Whereas Kimi 2.6, 1000B params is 0.7; GLM 5.1, 754B params is 1.0$.
Re:DeepSeek V4–almost on the frontier, a fraction of the price
by KronisLV on Saturday May 02, 2026 @05:52AM
(#47984945)
Here’s their pricing docs, they’re running a discount for now https://api-docs.deepseek.com/quick_start/pricing/
The 150M assumption of mine is for 100 USD at the regular prices (though even that needs sufficient cache hits). Anthropic subsidizes way more per-token I think, though.
Those tokens are heavily subsidized, but DeepSeek's API pricing is looking really good. For example, with an agentic coding setup (roughly 85% input, 15% output and around 90% cache reads) I'd get around 150M tokens per month for the same 100 USD. Even at more output tokens and worse cache performance, it'd still most likely be upwards of 100M.