Run Kimi-K2.7-Code locally: hardware requirements
Kimi-K2.7-Code needs about 622 GB of GPU memory for one person at 32K context, and about 107 GB more for each extra person at 64K. For 80 users working as AI agents, plan on 3413 GB, which takes 3 servers, each with 8 × B200.
Moonshot AI, China. From its files on Hugging Face, checked 2026-11-05.
Context per person: more context lets the model work with longer conversations and more files, and needs more GPU memory for every person. As a guide: 8K for short chat, 32K for chat over documents, 64K for AI agents, whose instructions, tools and files all count.
80 users means about 16 people working at the same moment. That share is a planning assumption; if you know yours, choose the total that matches it.
Work it out for your team: moonshotai/Kimi-K2.7-Code
On Hugging FaceMoonshot AI, China · 1.0T parameters · License: other
What you need for 16 active users
3 servers, each with 8 × B200
Each server runs its own copy of the model and holds 7 active users at 64K context.
Each server
3 TB RAM
64 CPU cores
2 TB NVMe
Rules of thumb
- GPU memory in total
- 3413 GB
- Weights, each server
- 595 GB (published files)
- Conversation cache, all active
- 1708 GB
- Engine (vLLM), each server
- 3.0 GB
Ways to need less
- Use 32K context per person2 servers, each with 6 × B200
- Use 8K context per person5 × B200
- Use a smaller model: gpt-oss-120b1 × B200
- Use a smaller model: Mixtral-8x7B-Instruct-v0.11 × B200
Your options by card
80 users means about 16 using the model at any moment. The others are not waiting; they are simply doing something else. If more become active than a build handles, they queue. 14 cards too small for this model are left out.
| Card | Memory | Active users, one server | For your 16 active |
|---|---|---|---|
| B200 | 180 GB | 7 | 3 servers, each with 8 × B200 |
| H200 NVL | 141 GB | 4 | 4 servers, each with 8 × H200 NVL |
| RTX PRO 6000 Blackwell | 96 GB | 1 | 16 servers, each with 8 × RTX PRO 6000 Blackwell |
These are memory limits. A card can hold more people than it can answer quickly, so test your own workload before you buy. More than one card in a server needs an engine that splits the model.
Kimi-K2.7-Code by team size
AI agents, 1 user in 5 active at a time, 64K context, as published.
| Total users | At once | GPU memory | What you need |
|---|---|---|---|
| 20 | 4 | 996 GB | 6 × B200 |
| 80 | 16 | 3413 GB | 3 servers, each with 8 × B200 |
| 120 | 24 | 4836 GB | 4 servers, each with 8 × B200 |
| 500 | 100 | 19200 GB | 15 servers, each with 8 × B200 |
Similar models
- Kimi-K2.5Moonshot AI · 1.0T · 2277 GB for 16 at once
- Kimi-K2.6Moonshot AI · 1.0T · 2277 GB for 16 at once
- Kimi-K2-InstructMoonshot AI · 1.0T · 2689 GB for 16 at once
- Kimi-K3Moonshot AI · 2.8T · 4067 GB for 16 at once
- Kimi-K3.1-ProMoonshot AI · 2.8T · 4067 GB for 16 at once
- Muse-Spark-Open-816B-A42BMeta · 822B · at least 1565 GB for 16 at once
How this is worked out
GPU memory = weights + a conversation cache for each person + the serving engine, from this model's files on Hugging Face. These are memory limits, not speed. The full method and its limits.
Questions
How much GPU memory does Kimi-K2.7-Code need?
Kimi-K2.7-Code needs about 622 GB of GPU memory for one person at 32K context, and about 107 GB more for each extra person at 64K. For 80 users working as AI agents, plan on 3413 GB, which takes 3 servers, each with 8 × B200.
How many active users can Kimi-K2.7-Code handle on one 96 GB GPU?
Kimi-K2.7-Code does not fit on one 96 GB card such as the RTX PRO 6000 Blackwell at 64K context. Compare larger cards or a smaller precision above.
Can Kimi-K2.7-Code run on a 24 GB GPU such as the RTX 4090?
Not as published for one person at 32K context: it needs about 622 GB. A 4-bit or 8-bit build may fit, if the model has one.
What license does Kimi-K2.7-Code use?
Hugging Face lists its license as other. Check the license on the model's page before you use it.
Running a model on your own servers?
NeueCode is private autonomous intelligence for enterprise: it takes on your organization's tasks on your own GPUs, with nothing leaving your network, a person approving every action, and all of it on record. It runs each model on one card, so read the single-card figures above.
What you need to run NeueCode