Thread

1. soerxpso · 1790547319 · focus · HN ↗

   I don't see why I would be interested in this model, considering the price difference. They advertise that it's the same as Kimi K3 in half the tokens. But the pricing is double the pricing of K3. So why do I care if it uses fewer tokens, if I'm paying double per token?

   1. verdverm · 1790548541 · focus · HN ↗

      the pricing is the same, where are you seeing double?

      1. tyingq · 1790557135 · focus · HN ↗

         K3 is cheaper from other providers. $1/$9, though I can't speak to how good those providers are.

         <a href="https:&#x2F;&#x2F;openrouter.ai&#x2F;moonshotai&#x2F;kimi-k3" rel="nofollow">https:&#x2F;&#x2F;openrouter.ai&#x2F;moonshotai&#x2F;kimi-k3

         1. verdverm · 1790558713 · focus · HN ↗

            as the saying goes, you get what you pay for

            we require ZDR and Fireworks provides that on contract, so for us they are the same price

            1. indigodaddy · 1790562385 · focus · HN ↗

               Check out Neuralwatt. They are ZDR and great energy based K3 pricing. They also have a K3-fast which is basically no reasoning (in addition to regular K3).

               1. verdverm · 1790564719 · focus · HN ↗

                  that is nothing like how we consume Ai

                  industry standard is price-per-1M tokens, don&#x27;t do something different, even Google caved and moved from their char based pricing to tokens (the fundamental unit of computation in ai)

                  GPUs are rented in $&#x2F;h, like every other piece of hardware in cloud

                  1. Bolwin · 1790576043 · focus · HN ↗

                     Why are you assuming the gpus are rented?

                     Anyway, they still have token based pricing if you prefer. The energy pricing is often cheaper though

                     1. verdverm · 1790605017 · focus · HN ↗

                        Not interested, it looks amateur

                        A trust page is required like <a href="https:&#x2F;&#x2F;trust.fireworks.ai" rel="nofollow">https:&#x2F;&#x2F;trust.fireworks.ai

                        Very happy with my $10&#x2F;month OpenCode Go sub for personal use

                        <a href="https:&#x2F;&#x2F;trust.opencode.ai&#x2F;">https:&#x2F;&#x2F;trust.opencode.ai&#x2F;

                  2. indigodaddy · 1790635601 · focus · HN ↗

                     Energy based is way more fundamental than token based. To each their own.
         2. polski-g · 1790567742 · focus · HN ↗

            K3 TOS says they must sell no lower than what Moonshot charges. If you have a provider selling for less than $15&#x2F;mtok, they are violating TOS from Moonshot.

            1. verdverm · 1790570248 · focus · HN ↗

               looks like the underlying vendor is charging correct prices <a href="https:&#x2F;&#x2F;inference.net&#x2F;models&#x2F;kimi-k3&#x2F;" rel="nofollow">https:&#x2F;&#x2F;inference.net&#x2F;models&#x2F;kimi-k3&#x2F;

               however the listing on open router has a `&#x2F;fp4` suffix, so perhaps this is an unlisted, quanted model for a lower price?

Unofficial Hacker News client; not affiliated with Y Combinator.