Apertus V1.5 8B — Hardware Requirements & GPU Compatibility
Vision
Apertus V1.5 8B is an 8.9-billion-parameter multimodal model from the Swiss AI Initiative, a collaboration between ETH Zurich, EPFL, and the Swiss National Supercomputing Centre. Extended from the text-only Apertus 1 through continued pretraining, it accepts images alongside text and can process spoken audio experimentally, letting it reason about visual content in a chat setting. Its compact size suits local deployment on a single consumer GPU once quantized. It is released under the Apache 2.0 license, allowing unrestricted commercial and research use, alongside a public acceptable-use policy governing deployment. Published in July 2026, Apertus V1.5 stands out for being fully open in a way few large models are: not just the weights but the training data, code, and development process are published, with an optional reasoning mode for harder prompts.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 8K)Best on 24 GB (Q4_K_M · ~111 tok/s · 32K)Macs from 16 GB (Q4_K_M · ~24 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Apertus V1.5 8B
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
vllm serve swiss-ai/Apertus-v1.5-8B --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3080 Ti (12 GB) (3 GB headroom) · Q4_K_M
- Generation speed
- ~101 tok/s
- generation speed
- Cost per 1M output tokens
- $0.34
- per 1M output tokens
How Much VRAM Does Apertus V1.5 8B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 4.2 GB | 9.8 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 32K context · ~158 tok/s | |
| 4.8 GB | 10.4 GB | 3-bit medium — 3-bit medium quantization | Fits: 32K context · ~137 tok/s | |
| 5.9 GB | 11.6 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 32K context · ~111 tok/s | |
| 7.0 GB | 12.7 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 32K context · ~94 tok/s | |
| 8.1 GB | 13.8 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 32K context · ~81 tok/s | |
| 9.8 GB | 15.5 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 32K context · ~67 tok/s | |
| 19.6 GB | 25.3 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: 16K context · ~33 tok/s |
Selected Q4_K_M
4-bit medium quantization8.5GBat 16K context
- Weights
- 5.2 GB
- KV cache
- 3.0 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 15.5 GB free · ~111 tok/s
≥50 tok/s needs ≥452 GB/s · RTX 5090 ~198, RTX 3090 Ti ~111
Which Hardware Can Run Apertus V1.5 8B?
Q4_K_M · 5.9 GBApertus V1.5 8B (Q4_K_M) requires 5.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 8+ GB is recommended. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3070 Ti. 151 devices with unified memory can run Apertus V1.5 8B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), NVIDIA Jetson Orin Nano 8GB (Super).
12 GB
12 GB tier →fits · 3.5 GB free · up to 32K ctx
10 GPUs · 1 device family
RTX 3080 Ti~101Simulate output on RTX 3080 TiRTX 5070~74Simulate output on RTX 5070RTX 4070~56Simulate output on RTX 4070iPhone~9Simulate output on Apple iPhone 17 Pro20 GB
fits · 11.5 GB free · up to 32K ctx
1 GPU
RX 7900 XT~82Simulate output on RX 7900 XT32–40 GB
fits · 23.5 GB free · up to 32K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~198Simulate output on RTX 5090M5 Max36 GB~55Simulate output on Mac Studio M5 Max (36 GB)M532 GB~18Simulate output on MacBook Air 13" M5 (32 GB)Orin~23Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~15Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to Apertus V1.5 8B
Apertus family · 5 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Granite Switch 4.1 8BIBM | — | 8.4 GB | ~111 | same size |
| Llm Jp 4.1 8B Thinkingllm-jp | — | 7.6 GB | ~115 | same size |
| Ornith 1.0 9BDeep Reinforce | — | 6.5 GB | ~106 | same size |
| LFM2.5 8B A1BLiquid AI | — | 5.6 GB | ~335 | same size |
| MiniCPM O 4 5OpenBMB | — | 8.3 GB | ~105 | same size |
| Chandradatalab-to | — | 8 GB | ~112 | same size |
Specifications
- Publisher
- swiss-ai
- Family
- Apertus
- Parameters
- 8.9B
- Release Date
- 2026-07-24
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Apertus V1.5 8B need?
Apertus V1.5 8B requires 5.9 GB of VRAM at Q4_K_M, or 19.6 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 8.9B × 4.8 bits ÷ 8 = 5.3 GB
KV Cache + Overhead ≈ 0.6 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
VRAM usage by quantization
What's the best quantization for Apertus V1.5 8B?
For Apertus V1.5 8B, Q4_K_M (5.9 GB) offers the best balance of quality and VRAM usage. Q5_K_M (7.0 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 4.2 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Apertus V1.5 8B on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Apertus V1.5 8B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 5.9 GB of usable unified memory (RAM minus macOS overhead).
Can I run Apertus V1.5 8B locally?
Yes — Apertus V1.5 8B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Apertus V1.5 8B?
At Q4_K_M, Apertus V1.5 8B can reach ~816 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~111 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 ÷ 5.88 × 0.65 = ~884 tok/s
Estimated speed at Q4_K_M (5.9 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Apertus V1.5 8B?
At Q4_K_M, the download is about 5.34 GB. The full-precision BF16 version is 17.81 GB. The smallest option (Q2_K) is 3.78 GB.
Which GPUs can run Apertus V1.5 8B?
52 consumer GPUs can run Apertus V1.5 8B at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT, AMD Radeon RX 7600. 40 GPUs have plenty of headroom for comfortable inference.
Which devices can run Apertus V1.5 8B?
151 devices with unified memory can run Apertus V1.5 8B at Q4_K_M (5.9 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (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.