Support Implicite/Explicite Prompt Caching
Many other providers offer prompt-caching with discounted pricing for cache hits, either implicitly (e.g., OpenAI, DeepSeek, Google, DeepInfra, NovitaAI, Fireworks) or explicitly (most notably Anthropic).
This capability can significantly reduce costs in agentic workflows, where a single session often re-sends the same context repeatedly (for example, when the model performs multiple tool calls in sequence and the shared conversation/context is included each time).
Today, Nebius Token Factory is at a cost disadvantage in these repeated, input-token-heavy scenarios compared to providers that support prompt caching and pass the savings through to customers.
Please add support for prompt caching (implicit or explicit), including discounted pricing for cached prompt tokens, to improve cost-efficiency for agentic and tool-using applications.
Log in to comment and vote
Comments13
Scramble Cloud
May 13
The newest OSS models are getting more and more expensive. Without Prompt Caching they are often unusable for specific cases like Agentic Coding, because its more expensive to use GLM 5.1 or DeepSeek V4 Pro by Nebius, then Sonnet 4.6 by Anthropic. I cannot see any more essential feature request then this.
Alex
Jun 20
I agree, although I like your models serving, it is literally cheaper to run something like Claude Sonnet 4.6 (even a normal chat with a lot of tool calls), than the open source models without caching with your pricing.
Nico Duldhardt
Jun 20
this is crazy that this is not supported yet. Get the EigenAI dudes to set that up as soon as possible
Stein Pleiter
Jun 23
Upvote, without this I cannot build what we need
ai ai
Jun 24
Much needed, not worth building on Nebius if this isn’t a feature. OpenAI en Anthropic models are still cheaper because of prompt caching.
Nils Czernig
Jun 26
Also for us a critical blocker for using the token factory
Ziaad
Jun 24
I would instantly switch to Nebius if this feature were to be available.
Currently a blocker for our company.
Michal Stefaniak
Jun 30
No prompt caching makes it economically unfeasible to run ANY agentic workflow/application.
Alexey Polyusov
Aug 25
Ребят, хотим кэширование.
No prompt caching makes it economically unfeasible to run ANY agentic workflow/application.
Дело говорит.
fenil modi
Sep 2
common its 2026
still no catch token
how is this neo cloud