K2 Horizon 7B — VRAM, speed & local setup
Written by Jakub Rusinowski · Last updated
The best independently scored model of its size when checked: 21 on the Artificial Analysis Intelligence Index, against 11 for Qwen3.5-9B. A 7B dense core plus a 250K-token vocabulary, 9.0B parameters stored, so the official Q4_K_M GGUF is 5.6 GB; with a 16K context it needs about 8.8 GB, so a 12 GB card is the practical minimum. Its makers report SWE-bench Verified 70.6 and Terminal-Bench 2.1 39.1, against 50.8 and 29.2 for Qwen3.5-9B on the same card — vendor-reported, so the independent score is the one to lean on. Apache 2.0, 512K native context.
K2 Horizon 7B needs about 6 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.
VRAM and speed by quantization
Quoted against NVIDIA RTX 4090 (24 GB). Includes the KV cache at 8K context, so it reads higher than the headline figure.
| Quant | Memory | VRAM | Speed (est.) | Fit |
|---|---|---|---|---|
| Q2_K 2.63 bpw | 5 GB | ~144 tok/s | Fits | |
| Q3_K_M 3.41 bpw | 5.8 GB | ~123 tok/s | Fits | |
| Q4_K_M 4.83 bpw | 7.4 GB | ~98 tok/s | Fits | |
| Q5_K_M 5.67 bpw | 8.4 GB | ~87 tok/s | Fits | |
| Q6_K 6.56 bpw | 9.4 GB | ~78 tok/s | Fits | |
| Q8_0 8.50 bpw | 11.6 GB | ~64 tok/s | Fits | |
| F16 16.00 bpw | 20 GB | ~37 tok/s | Fits |
Black marker = usable memory on the NVIDIA RTX 4090 (24 GB). Estimates from the memory-bandwidth roofline on the methodology page. K2 Horizon 7B VRAM calculator →
Get K2 Horizon 7B running
The cheapest catalogued GPU that runs K2 Horizon 7B is the Intel Arc B570 (10 GB).
or compare on Vast.ai from $0.35/hr (typical low · varies)
As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.
How to run K2 Horizon 7B
No Ollama library tag yet. Run the official GGUF with llama.cpp: the files declare the k2-horizon architecture, which mainline llama.cpp supports; an older Ollama or LM Studio build may not load it. The model card's own serving route is vLLM or SGLang.
Specifications
Verified — Checked against the primary source — the model card or the vendor spec page — and corroborated by a second independent source.
- Parameters
- 9.0 Billion (7B core + embeddings)
- Context window
- 524,288
- Architecture
- Dense Transformer (GQA)
- Provider
- MBZUAI Institute of Foundation Models
- Licence
- Apache 2.0
- Specified at
- Q4_K_M
- System RAM
- 16 GB
- Record updated
- 2026-10-08
Commercial use permitted. No usage restrictions beyond attribution.
Quality and use cases
Scores as published by the model’s authors or an independent evaluator — quality, not throughput, and not measured by us.
| Benchmark | Score | Provenance |
|---|---|---|
| SWE-bench Verified | 70.6 / 100 % | reported · https://huggingface.co/IFM/K2-Horizon-7B |
| Terminal-Bench 2.1 | 39.1 / 100 % | reported · https://huggingface.co/IFM/K2-Horizon-7B |
Other K2 Horizon sizes
K2 Horizon 7B — frequently asked questions
How much VRAM does K2 Horizon 7B need?
About 6 GB at Q4_K_M — quantized weights plus framework overhead, before any KV cache. The cache grows with context length and is added on top; the table above folds it in. Apple Silicon counts unified memory toward the same figure.
Does K2 Horizon 7B run on an RTX 4090 (24 GB)?
Yes. K2 Horizon 7B needs about 6 GB at Q4_K_M, inside a 24 GB card, at an estimated 98 tokens/sec.
How do I run K2 Horizon 7B locally?
No Ollama library tag yet. Run the official GGUF with llama.cpp: the files declare the k2-horizon architecture, which mainline llama.cpp supports; an older Ollama or LM Studio build may not load it. The model card's own serving route is vLLM or SGLang. Running the published tag would send your prompts to a hosted GPU rather than your own machine.
What other sizes does K2 Horizon come in?
K2 Horizon 3.7B (4 GB), K2 Horizon 7B (6 GB), K2 Horizon MoVA 36B-A4B (23 GB). Every size shares the family's training and licence; the larger ones score higher and need proportionally more memory.