Serving Kimi K2.7 Code with Triton Inference Server

Kimi K2.7 Code has 1000 billion parameters. Triton Inference Server targets industrial deployment across several models. This page gives the memory required and the platforms that qualify.

By QDNA · Published on 27 July 2026 · Updated on 15 September 2026

Short answer. Triton Inference Server serves Kimi K2.7 Code as soon as the machine offers at least 575 GB of memory, the weight footprint in NVFP4. 6 of the 8 platforms in the catalogue meet that bar.

What is the minimum memory?

Kimi K2.7 Code totals 1000 billion parameters. In NVFP4 its weights take about 575 GB including the runtime margin: 1,000 billion parameters × 0.5 byte (NVFP4) = 500.0 GB of weights; × 1.15 runtime margin = 575.0 GB. In FP8 the footprint doubles, to about 1,150 GB (1,000 billion parameters × 1 byte (FP8) = 1,000.0 GB of weights; × 1.15 runtime margin = 1,150.0 GB.) The attention cache is not calculated on this page. These models’ architectures (compressed latent, linear or sparse attention, Mamba layers) share no per-token formula; it is measured on the machine, with the target context and concurrency [TO BE MEASURED].

What is Triton Inference Server for?

Triton Inference Server targets industrial deployment across several models. See the Triton Inference Server and Kimi K2.7 Code fact sheets.

On which platforms?

How much memory for Kimi K2.7 Code with Triton Inference Server?

About 575 GB in NVFP4 for the weights, excluding the attention cache.

Does Triton Inference Server suit Kimi K2.7 Code?

Triton Inference Server serves Kimi K2.7 Code as soon as the machine offers at least 575 GB of memory, the weight footprint in NVFP4. 6 of the 8 platforms in the catalogue meet that bar.

Method and sources

Memory footprints are calculated, not measured: a reading on real hardware may differ depending on the engine and the exact weight format. Every figure below carries its source and the date of its last check.

Kimi K2.7 Code: 1,000 billion parameters, 32 billion active, 262,144-token context, Modified MIT licence. Hugging Face repository (config.json, API, model card), checked on 2 September 2026. 1 trillion announced (model card), 8 of 384 experts active; the repository holds 1,027 billion elements, published in 4-bit (595 GB).
Bytes per parameter: NVFP4 0.5 (4 bits per parameter, hence 0.5 byte; 4-bit published weights weigh 0.54 to 0.56 byte per parameter with their scales (DeepSeek V4 Pro 865 GB for 1,599 billion, Kimi K3 1,561 GB for 2,780 billion)); FP8 1 (8 bits per parameter (E4M3), hence 1 byte, block scales not counted). Source: connaissance/faits.yaml, quantifications family, checked on 2026-08-31.
Runtime margin × 1.15: QDNA operating assumption, not measured: 15% above the weights for activations, buffers and fragmentation.
“Fits” threshold at 75% of memory: QDNA assumption, not measured, which keeps the remaining quarter for the attention cache and concurrency.