The best GPU for running Kimi K3 (MoE) (2026)
Ranked from live rental prices and computed VRAM fit, not opinion. Every card is judged at MXFP4, the format Kimi K3 (MoE)ships in (its 4-bit builds are no smaller, so it's also what people run) — at 8k context, so it's a fair comparison. A card that can't fit Kimi K3 (MoE)at that quality isn't listed, because it could only run a crushed version.
No single card in our database fits Kimi K3 (MoE) at its own MXFP4 release and 8k context — at 2780B parameters it needs more memory than any one consumer or workstation GPU has, so this is a multi-GPU or datacenter decision.
See one fast card vs two big ones and the cheapest way to serve a 70B, or the full breakdown on the cost-to-run page.
Check it yourself
See exactly what fits, at any context length
Running Kimi K3 (MoE): common questions
How much VRAM do you need to run Kimi K3 (MoE)?
Can an RTX 4090 run Kimi K3 (MoE)?
Can an RTX 3090 run Kimi K3 (MoE)?
Can an RTX 4060 Ti 16GB run Kimi K3 (MoE)?
Before you buy
How do you get a Kimi K3 API key, and what does it cost per 1M tokens?
The Kimi API is pay as you go from Moonshot's own platform, and Kimi K3 unlocks after a $1 top-up; we found no free tier, only a $5 voucher once you have added $5. Here are Moonshot's real per-million-token prices, the same-model price on Novita AI, and why running K3 yourself is not a realistic option.
Kimi K3's open weights are 1.56 TB. Can your GPU actually run it?
Moonshot's Kimi K3 is open to download. It is also 2.8 trillion parameters in 1.56 TB of MXFP4 files, and the smallest community quant is 509 GB. The real memory math, what vLLM says you need, and what to run at home instead.
How much VRAM do you need?
About 0.57 GB per billion parameters at 4-bit, plus room for context: the rule behind the numbers on this page, with worked examples.
The best GPU for local AI right now
The card with enough VRAM for your model at the lowest price, budget by budget.