How much VRAM Kimi K3 (MoE) needs, and what it costs to run
Some links on this page are referral links: we may earn a commission, at no cost to you.
Trying to decide what to buy? The best GPU for Kimi K3 (MoE) →
Just want to use it? Per-token API
$3.00in·$15.00out
per 1M tokens · Kimi K3 on Novita AI · read Oct 29, 2026
No GPU to manage: you pay per token, which suits occasional use. Your prompts run on Novita's servers, not your own machine.
Price as listed on Novita's pricing page on Oct 29, 2026; it may have changed.
Memory needed, by quantisation
At 16k context. Weights + KV cache + ~0.8 GB overhead. Kimi K3 (MoE) ships in MXFP4, so its own release leads and no bigger copy is listed: it would only store the same numbers in more bytes. Memory is in binary GB (1,024³ bytes), the unit GPU memory is sold in; Hugging Face and Ollama list downloads in decimal GB, about 7% more.
| Quant | Weights | KV cache | Total | Quality |
|---|---|---|---|---|
| MXFP4official | 1453.7 GB | 0.4 GB | 1454.9 GB | The model's own release (MXFP4, a 1560.9 GB download): the best quality there is. A bigger copy only stores the same numbers in more bytes. Its 4-bit GGUF builds keep the 4-bit experts as they are, so they come out about this size too. |
How these numbers are produced
- Model shape comes from the model's own config.json — 24 layers, 96 attention heads, 1 KV heads.
- Weight size is parameters × effective bits-per-weight. Those constants are checked against real published quantised file sizes — median error 0.7% across 40 measurements.
- KV cache is 2 × layers × KV-heads × head-dim × context × 2 bytes. Using KV-heads rather than attention heads is what makes this correct for grouped-query attention; treating a GQA model as multi-head overstates the cache by up to 8×.
- Tokens/sec is an estimate, not a benchmark. Generation is memory-bandwidth-bound, so this is bandwidth ÷ weight-bytes derated to 75%. Real throughput depends on your runtime, batch size and kernels. We don't run our own hardware tests — see our editorial policy.
- This is a mixture-of-experts model. Generation reads only the experts active for each token (16 of 896), so throughput is scored on ~104.0B active parameters, not the full 2780B. Memory is the opposite: every expert must still be resident in VRAM, so the figures above use the full weight set.Hybrid attention: 93 layers = 69 KDA (Kimi Delta Attention, linear: a fixed-size state, not a growing cache) + 24 gated MLA full-attention layers (model card). `layers`/kv_heads/head_dim = the 24 MLA layers: kv_lora_rank 512 + qk_rope_head_dim 64 = 576 cached numbers per token per layer, mapped as kv_heads 1 x head_dim 288 x 2 = 576 (same convention as GLM-5.3). KDA state is fixed, about 69 layers x 96 heads x 128x128 = ~0.2 GB per sequence at bf16, not counted. Active = vendor 104B; params = HF safetensors total (~2.78T; card says 2.8T), routed experts ship in MXFP4 (~1.56 TB of files); includes the vision encoder. Release date = first public commit on HF (27 Jul 2026; the repo itself was created 13 Jun). HF config.json + model card read 29 Oct 2026.
- Rental prices are pulled hourly from provider APIs (last updated Thu, 12 Nov 2026 22:20:09 GMT). Spot/interruptible pricing can change or vanish without notice.
Try it interactively
Drag the context slider and watch the KV cache fill the card
Worth reading 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.