Kimi Dev 72B — Hardware Requirements & GPU Compatibility
ChatCode
Kimi Dev 72B is Moonshot AI's developer-focused model built on the Qwen2.5-72B architecture, specifically optimized for coding tasks, tool use, and agentic workflows. It combines strong general-purpose chat abilities with specialized developer capabilities, making it a compelling choice for software engineering assistance. At 72 billion parameters it requires substantial hardware, typically needing 40+ GB of VRAM at 4-bit quantization, which puts it in reach of dual consumer GPU setups or single professional cards like the A100 or RTX 6000 Ada. If you are primarily looking for a local coding assistant with strong reasoning skills, Kimi Dev is a top-tier option in the 70B class.
2.9K downloads
384 likesSpecifications
- Publisher
- Moonshot AI
- Family
- Kimi
- Parameters
- 72B
- Architecture
- Qwen2ForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 152,064
- Release Date
- 2025-06-17
- License
- MIT
Get Started
HuggingFace
How Much VRAM Does Kimi Dev 72B Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 145.0 GB | 187.3 GB | 144.00 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Kimi Dev 72B?
BF16 · 145.0 GBKimi Dev 72B (BF16) requires 145.0 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 189+ GB is recommended. Using the full 131K context window can add up to 42.3 GB, bringing total usage to 187.3 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Kimi Dev 72B?
BF16 · 145.0 GB4 devices with unified memory can run Kimi Dev 72B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Pro M2 Ultra (192 GB).
Runs great
— Plenty of headroom
Decent
— Enough memory, may be tightRelated Models
Frequently Asked Questions
- How much VRAM does Kimi Dev 72B need?
Kimi Dev 72B requires 145.0 GB of VRAM at BF16. Full 131K context adds up to 42.3 GB (187.3 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 72B × 16 bits ÷ 8 = 144 GB
KV Cache + Overhead ≈ 1 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 43.3 GB (at full 131K context)
VRAM usage by quantization
BF16145.0 GBBF16 + full context187.3 GB- Can NVIDIA GeForce RTX 5090 run Kimi Dev 72B?
No — Kimi Dev 72B requires at least 145.0 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run Kimi Dev 72B on a Mac?
Kimi Dev 72B requires at least 145.0 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Kimi Dev 72B locally?
Yes — Kimi Dev 72B can run locally on consumer hardware. At BF16 quantization it needs 145.0 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Kimi Dev 72B?
At BF16, Kimi Dev 72B can reach ~20 tok/s on AMD Instinct MI300X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 145.0 × 0.55 = ~20 tok/s
Estimated speed at BF16 (145.0 GB)
~20 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Kimi Dev 72B?
At BF16, the download is about 144.00 GB.
- Which GPUs can run Kimi Dev 72B?
No single consumer GPU has enough VRAM to run Kimi Dev 72B at BF16 (145.0 GB). Multi-GPU or professional hardware is required.
- Which devices can run Kimi Dev 72B?
4 devices with unified memory can run Kimi Dev 72B at BF16 (145.0 GB), including Mac Pro M2 Ultra (192 GB), Mac Studio M2 Ultra (192 GB), NVIDIA DGX A100 640GB, NVIDIA DGX H100. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.