Kimi Linear 48B A3B Base — Hardware Requirements & GPU Compatibility
Chat
Kimi Linear 48B A3B Base is a 49.1B-parameter open language model from Moonshot AI in the Kimi family. At Q4_K_M it needs about 30.28 GB of VRAM — see which GPUs and Macs can run it below.
124.6K downloads
84 likesSpecifications
- Publisher
- Moonshot AI
- Family
- Kimi
- Parameters
- 49.1B
- Architecture
- KimiLinearForCausalLM
- Vocabulary Size
- 163,840
- Release Date
- 2025-10-30
- License
- MIT
Get Started
HuggingFace
How Much VRAM Does Kimi Linear 48B A3B Base Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 21.7 GB | — | 20.88 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 22.3 GB | — | 21.49 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 24.8 GB | — | 23.95 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 25.4 GB | — | 24.56 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 30.3 GB | — | 29.47 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 35.8 GB | — | 35.00 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 41.3 GB | — | 40.53 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 49.9 GB | — | 49.12 GB | 8-bit quantization, near-lossless |
est.= calculated VRAM estimate; no published GGUF file found for that quantization yet. Other rows are verified against real community uploads.
Which GPUs Can Run Kimi Linear 48B A3B Base?
Q4_K_M · 30.3 GBKimi Linear 48B A3B Base (Q4_K_M) requires 30.3 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 40+ GB is recommended. 1 GPU can run it, including NVIDIA GeForce RTX 5090.
All compatible consumer-level GPUs are running near their VRAM limit. You may also want to consider professional GPUs (e.g., NVIDIA A100, H100) which offer significantly more VRAM. For more headroom and better throughput, consider a multi-GPU configuration with tensor parallelism (supported by tools like vLLM, llama.cpp, or text-generation-inference).
Decent
— Enough VRAM, may be tightWhich Devices Can Run Kimi Linear 48B A3B Base?
Q4_K_M · 30.3 GB31 devices with unified memory can run Kimi Linear 48B A3B Base, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Studio M4 Max (36 GB).
Runs great
— Plenty of headroomWhere to Download Kimi Linear 48B A3B Base
Community quantizations of this model — GGUF for llama.cpp, Ollama, and LM Studio, plus AWQ/MLX variants where available.
Related Models
Frequently Asked Questions
- How much VRAM does Kimi Linear 48B A3B Base need?
Kimi Linear 48B A3B Base requires 30.3 GB of VRAM at Q4_K_M, or 99.0 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 49.1B × 4.8 bits ÷ 8 = 29.5 GB
KV Cache + Overhead ≈ 0.8 GB (at 2K context + ~0.3 GB framework)
VRAM usage by quantization
Q4_K_M30.3 GB- Can NVIDIA GeForce RTX 4090 run Kimi Linear 48B A3B Base?
Yes, at IQ3_M (22.9 GB) or lower. Higher quantizations like Q3_K_M (24.8 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
- What's the best quantization for Kimi Linear 48B A3B Base?
For Kimi Linear 48B A3B Base, Q4_K_M (30.3 GB) offers the best balance of quality and VRAM usage. Q5_K_S (34.6 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 14.3 GB.
VRAM requirement by quantization
IQ2_XXS14.3 GBIQ3_XS21.1 GBQ3_K_M24.8 GBQ4_K_M ★30.3 GBQ5_K_S34.6 GBBF1699.0 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Kimi Linear 48B A3B Base on a Mac?
Kimi Linear 48B A3B Base requires at least 14.3 GB at IQ2_XXS, 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 Linear 48B A3B Base locally?
Yes — Kimi Linear 48B A3B Base can run locally on consumer hardware. At Q4_K_M quantization it needs 30.3 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Kimi Linear 48B A3B Base?
At Q4_K_M, Kimi Linear 48B A3B Base can reach ~171 tok/s on AMD Instinct MI350X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 30.3 × 0.65 = ~527 tok/s
Estimated speed at Q4_K_M (30.3 GB)
~527 tok/s~527 tok/s~463 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Kimi Linear 48B A3B Base?
At Q4_K_M, the download is about 29.47 GB. The full-precision BF16 version is 98.25 GB. The smallest option (IQ2_XXS) is 13.51 GB.
- Which GPUs can run Kimi Linear 48B A3B Base?
1 consumer GPU can run Kimi Linear 48B A3B Base at Q4_K_M (30.3 GB). Top options include NVIDIA GeForce RTX 5090.
- Which devices can run Kimi Linear 48B A3B Base?
35 devices with unified memory can run Kimi Linear 48B A3B Base at Q4_K_M (30.3 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB), Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.