// Kimi · Text LLM
Kimi-K2-Thinking: VRAM & local deployment
1T total · 32B active (MoE) parameters. About 668 GB at 4-bit and 2.20 TB at BF16 with one 8K-token request.
released 2025-11-04licence: Commercial with conditionsmemory figures are estimates
01Specs
| Publisher | Moonshot AI · moonshotai |
|---|---|
| Family / series | Kimi · Kimi-K2-Thinking |
| Type | Text LLM (text-generation) |
| Parameters | 1T total · 32B active (MoE) Source: official model card“Total Parameters 1T · Activated Parameters 32B” |
| Context length | 256K tokens (262,144)max_position_embeddings in config.json |
| Architecture | 61 layers · MLA · MoE |
| Released weights | packed quantised tensors (I32, e.g. MXFP4 / INT4) · 594 GBdownload |
| Licence | modified-mit · Commercial with conditionsCommercial use allowed under the publisher’s own licence terms (user caps, acceptable-use policy, attribution or region limits). Read it before shipping. |
| Released | 2025-11-04 (Hugging Face repo created) |
02How much VRAM it needs (estimate)
Total = weights + 10% + 0.5GB overhead + KV cache for one request (FP16 KV). More concurrent users multiply the KV column.
| Precision | Weights | Total @ 8K | Total @ 32K | Total @ 128K | Smallest GPU | Mac |
|---|---|---|---|---|---|---|
| BF16 / FP16 (original) | 2.00 TB | 2.20 TB | 2.20 TB | 2.21 TB | 2 × 8 × B300 nodes (4,608GB) | — |
| FP8 / INT8 / Q8_0 | 1.06 TB | 1.17 TB | 1.17 TB | 1.18 TB | 8 × B200 (1,440GB) | — |
| 4-bit (Q4_K_M / AWQ / GPTQ-Int4) | 606 GB | 668 GB | 670 GB | 677 GB | 8 × H200 (1,128GB) | — |
KV cache per request: 8K 0.6 GB · 32K 2.3 GB · 128K 9.2 GB
Bits per weight: BF16 16, FP8 8.5, 4-bit 4.85. “Fits” = total ≤ 90% of memory; Mac GPU can use ~2/3 of unified memory (≤36GB) or ~3/4 (larger). For MoE, all experts must be in memory, so size by total parameters. Method →
03How to run it
Only runtimes the publisher or the runtime’s official library names for this model.
- vLLM — mentioned in the official model card
- SGLang — mentioned in the official model card
- Transformers — mentioned in the official model card
- KTransformers — mentioned in the official model card
04Official repositories & quantised builds
1 official repo(s) from moonshotai.
- Kimi-K2-Thinking main
05Other Kimi models
All 14 →| Model | Params | 4-bit total | Released |
|---|---|---|---|
| Kimi-K3 | 2.8T / 104B | 1.87 TB | 2026-06-13 |
| Kimi-K2.7-Code | 1T / 32B | 668 GB | 2026-06-11 |
| Kimi-K2.6 | 1T / 32B | 668 GB | 2026-04-14 |
| Kimi-K2.5 | 1T / 32B | 668 GB | 2026-01-01 |
| Kimi-Linear-48B-A3B-Instruct | 49.1B / 3B | 33.3 GB | 2025-10-30 |
| Kimi-K2-Instruct-0905 | 1T / 32B | 668 GB | 2025-09-03 |
| Kimi-K2-Instruct | 1T / 32B | 668 GB | 2025-07-11 |
| Kimi-VL-A3B-Thinking-2506 | 16.4B / 3B | 11.7 GB | 2025-06-21 |
06Sources
Data checked 2026-01-30. Parameters, context and licence are copied from the official repository; memory figures are our estimates.