K2 Horizon 3.7B — Hardware Requirements & GPU Compatibility
Chat
K2 Horizon 3.7B is a 5.1B-parameter open language model from IFM. It supports a context window of up to 524,288 tokens. At Q4_K_M it needs about 3.52 GB of VRAM — see which GPUs and Macs can run it below.
1.3K downloads
13 likesSpecifications
- Publisher
- IFM
- Parameters
- 5.1B
- Architecture
- K2HorizonForCausalLM
- Context Length
- 524,288 tokens
- Vocabulary Size
- 250,624
- Release Date
- 2026-09-01
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does K2 Horizon 3.7B Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 2.6 GB | 50.8 GB | 2.15 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 2.7 GB | 50.8 GB | 2.21 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 3.0 GB | 51.1 GB | 2.47 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 3.0 GB | 51.1 GB | 2.53 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 3.5 GB | 51.6 GB | 3.03 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 4.1 GB | 52.2 GB | 3.60 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 4.7 GB | 52.8 GB | 4.17 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 5.5 GB | 53.7 GB | 5.06 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run K2 Horizon 3.7B?
Q4_K_M · 3.5 GBK2 Horizon 3.7B (Q4_K_M) requires 3.5 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 5+ GB is recommended. Using the full 524K context window can add up to 48.1 GB, bringing total usage to 51.6 GB. 50 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run K2 Horizon 3.7B?
Q4_K_M · 3.5 GB59 devices with unified memory can run K2 Horizon 3.7B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, iPhone 17.
Runs great
— Plenty of headroom
Decent
— Enough memory, may be tightWhere to Download K2 Horizon 3.7B
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 K2 Horizon 3.7B need?
K2 Horizon 3.7B requires 3.5 GB of VRAM at Q4_K_M, or 10.6 GB at BF16. Full 524K context adds up to 48.1 GB (51.6 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 5.1B × 4.8 bits ÷ 8 = 3 GB
KV Cache + Overhead ≈ 0.5 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 48.7 GB (at full 524K context)
VRAM usage by quantization
Q4_K_M3.5 GBQ4_K_M + full context51.6 GB- What's the best quantization for K2 Horizon 3.7B?
For K2 Horizon 3.7B, Q4_K_M (3.5 GB) offers the best balance of quality and VRAM usage. Q5_0 (3.6 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 1.9 GB.
VRAM requirement by quantization
IQ2_XXS1.9 GBIQ3_XS2.6 GBQ3_K_L3.1 GBQ4_K_M ★3.5 GBQ5_03.6 GBBF1610.6 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run K2 Horizon 3.7B on a Mac?
K2 Horizon 3.7B requires at least 1.9 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 K2 Horizon 3.7B locally?
Yes — K2 Horizon 3.7B can run locally on consumer hardware. At Q4_K_M quantization it needs 3.5 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is K2 Horizon 3.7B?
At Q4_K_M, K2 Horizon 3.7B can reach ~1364 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~186 tok/s. 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 ÷ 3.5 × 0.65 = ~1477 tok/s
Estimated speed at Q4_K_M (3.5 GB)
~1477 tok/s~186 tok/s~1477 tok/s~1364 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of K2 Horizon 3.7B?
At Q4_K_M, the download is about 3.03 GB. The full-precision BF16 version is 10.12 GB. The smallest option (IQ2_XXS) is 1.39 GB.
- Which GPUs can run K2 Horizon 3.7B?
50 consumer GPUs can run K2 Horizon 3.7B at Q4_K_M (3.5 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 50 GPUs have plenty of headroom for comfortable inference.
- Which devices can run K2 Horizon 3.7B?
59 devices with unified memory can run K2 Horizon 3.7B at Q4_K_M (3.5 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Apple iPhone 17 Pro, Asus ROG Flow Z13 (2025, 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.