How much VRAM does K2 Horizon 7B need?

About 6.5 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 10.8 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.

Will K2 Horizon 7B run on a 8GB GPU?

Yes — at Q4_K_M and 4K context it needs about 6.5 GB, which leaves 1.5 GB spare on a RTX 4060, or on any of the other 8 GB cards here (9 in all). That is the smallest size in this index that clears it comfortably; 76 of 78 do.

Which quantization of K2 Horizon 7B should I use?

No published quality comparison exists for this model, so pick by footprint: Q4_K_M is the level most people run, and a higher bits-per-weight level is more faithful. All 4 levels in the index are Q4_K_M, Q5_K_M, Q6_K, Q8_0.

K2 Horizon 7B

MBZUAI K2 Horizon

The small dense member of MBZUAI IFM's K2 Horizon family, Apache 2.0, released with its training data. It is called 7B but holds 9.0B parameters: the 250K-token vocabulary's input and output tables add about 2B. Q4_K_M is about 6.5 GB at 4K, comfortable on an 8 GB card at that length; a 16 GB card holds it comfortably to 32K (10.8 GB), and 128K needs about 25.6 GB because every layer keeps an ordinary KV cache. llama.cpp runs it as `k2-horizon`. Sizes come from IFM's own GGUF files (bartowski's Q4_K_M is about 0.2 GB larger); no per-level quality loss has been published.

⬇ 35.6K HF downloads♥ 79 likesIFM/K2-Horizon-7B-GGUF· stats from 10/21/2026

Consumer GPUMac / Apple SiliconCPU / VPS

524K

Max Context

Quant Variants

GGUF Q8_0

Best Quality

—

Accuracy Retained

Calculate VRAM Hugging Face Compare

Quantization Variants

Per-quant VRAM, quality loss, and inference speed on RTX 4090

Measured = run on this site · Estimated = not run here: calculated, or a figure this site has not verified · Community = public reports

Format	Level	BPW	VRAM	PPL Loss	Speed	Source	Actions
GGUF	Q4_K_M	4.97	6.5 GB	—	—	Estimated	Calc HF
GGUF	Q5_K_M	5.74	7.4 GB	—	—	Estimated	Calc HF
GGUF	Q6_K	6.57	8.3 GB	—	—	Estimated	Calc HF
GGUF	Q8_0	8.5	10.6 GB	—	—	Estimated	Calc HF

Which file each size is for

A file size is not a memory figure: the VRAM column adds the KV cache at 4K context and runtime buffers to the weights. A vision projector (mmproj), where the repo ships one, is a separate download and is in neither number.

GGUF Q4_K_MIFM/K2-Horizon-7B-GGUF · K2-Horizon-7B-Q4_K_M.gguf · 5.59 GB (5,592,217,984 bytes) · listing read 2026-10-21
GGUF Q5_K_MIFM/K2-Horizon-7B-GGUF · K2-Horizon-7B-Q5_K_M.gguf · 6.47 GB (6,466,075,008 bytes) · listing read 2026-10-21
GGUF Q6_KIFM/K2-Horizon-7B-GGUF · K2-Horizon-7B-Q6_K.gguf · 7.39 GB (7,394,548,096 bytes) · listing read 2026-10-21
GGUF Q8_0IFM/K2-Horizon-7B-GGUF · K2-Horizon-7B-Q8_0.gguf · 9.57 GB (9,573,964,160 bytes) · listing read 2026-10-21

Similar models

Compare with K2 Horizon

36B-A4B

K2 Horizon MoVA 36B-A4B

MBZUAI K2 Horizon

Consumer GPUPro GPU

24.2 GBmin VRAM—accuracy

The mixture-of-experts member of MBZUAI IFM's K2 Horizon family: 37.5B parameters with about 4B active per token, Apache 2.0, released with its training data. Every one of its 48 layers keeps an ordinary KV cache, so context is not cheap: Q4_K_M is about 24.2 GB at 4K, which is tight on a 24 GB card, and comfortable on a 32 GB card up to 16K. The 512K window would need about 129 GB. llama.cpp runs it as `k2-horizon`. Sizes come from IFM's own GGUF files (bartowski's Q4_K_M is about 0.3 GB larger); no per-level quality loss has been published.

32B

K2 Horizon 32B

MBZUAI K2 Horizon

Consumer GPUPro GPU

23.1 GBmin VRAM—accuracy

The dense 32B member of MBZUAI IFM's K2 Horizon family (34.8B counted, Apache 2.0, released with its training data). All 64 layers keep an ordinary KV cache with 8 KV heads: Q4_K_M is about 23.1 GB at 4K, tight on a 24 GB card and comfortable on a 32 GB card up to 16K; each extra 32K of context adds roughly 8.8 GB. llama.cpp runs it as `k2-horizon`. Sizes come from IFM's own GGUF files (bartowski's Q4_K_M is about 1 GB larger); no per-level quality loss has been published.

Qwen3-VL 8B Instruct

Alibaba Qwen3-VL

Consumer GPUMac / Apple Silicon

5.3 GBmin VRAM99.7%accuracy

Current-generation vision-language model that still fits a single 8–12GB card at Q4 (~5.9GB). The realistic multimodal option for people without a 24GB GPU — note the vision encoder adds VRAM the KV-cache math below does not model.

Qwen3.5 9B

Alibaba Qwen3.5

Consumer GPUMac / Apple Silicon

6.6 GBmin VRAM—accuracy

The newest small Qwen and a common pick for 8 to 16 GB cards. Only 8 of its 32 layers keep a growing KV cache, so it can use a long context on modest hardware. At Q4_K_M it needs about 6.6 GB at 4K and 7.5 GB at 32K, comfortable on an 8 GB card. Its full 256K window needs about 15.2 GB. Q8_0 needs about 10.4 GB, comfortable from 12 GB up. Reasoning is off by default on the small Qwen3.5 sizes, and it reads images through a separate vision encoder. Sizes come from the published GGUF file sizes, and no per-level quality loss has been published.

Running K2 Horizon 7B locally

At Q4_K_M and 4K of context, K2 Horizon 7B needs about 6.5 GB — 5.3 GB of weights, 0.56 GB of KV cache and a 0.6 GB activation buffer. The smallest cards in this index that clear that comfortably are 8 GB ones — 9 of them, from the RTX 4060 up — and 76 of the 78 cards here do. These are calculated figures, not measurements: the estimate stops counting a card as comfortable at 88% of its VRAM, which is roughly the room a desktop session needs.

What longer context costs

Going from 4K to 32K adds about 3.94 GB, taking the total to 10.8 GB. Weights do not move with context — only the KV cache does, and it grows linearly, so this is the number to watch when planning for long documents. The model's native window is 512K; holding all of it at Q4_K_M would need about 85 GB.

Which build to download

This index only tracks K2 Horizon 7B in GGUF, across 4 levels (Q4_K_M, Q5_K_M, Q6_K, Q8_0). No per-level perplexity sweep has been published for this model, so the quality column is empty rather than estimated — within one model, more bits per weight is the only ordering the data supports. No throughput has been measured for this model on this site's hardware. GGUF runs on llama.cpp and Ollama across NVIDIA, AMD and Apple silicon; AWQ and GPTQ target vLLM on CUDA and ROCm; EXL2 is ExLlamaV2 — CUDA only, and now archived, so local use rather than serving.

Common questions

How much VRAM does K2 Horizon 7B need?: About 6.5 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 10.8 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.
Will K2 Horizon 7B run on a 8GB GPU?: Yes — at Q4_K_M and 4K context it needs about 6.5 GB, which leaves 1.5 GB spare on a RTX 4060, or on any of the other 8 GB cards here (9 in all). That is the smallest size in this index that clears it comfortably; 76 of 78 do.
Which quantization of K2 Horizon 7B should I use?: No published quality comparison exists for this model, so pick by footprint: Q4_K_M is the level most people run, and a higher bits-per-weight level is more faithful. All 4 levels in the index are Q4_K_M, Q5_K_M, Q6_K, Q8_0.

Where this model fits

Sized at Q4_K_M with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory.

Runs comfortably on: RTX 40608GB · +1.5 RTX 4060 Ti 8G8GB · +1.5 Radeon RX 9060 XT 8G8GB · +1.5 RTX 5060 Ti 8G8GB · +1.5 RTX 50608GB · +1.5
Just misses: Mac M1 8G+0.5GB over Mac M3 8G+0.5GB over
Ships in: GGUF
Try it yourself: Size it in the calculator Generate the run command

How to actually run this

Deployment guides for this model and this class of hardware.

This page's figures change when the model or the runtime does.Last updated 2026-10-21 RSS → /feed.xml