llmrun
HcompanyQwen3_5MoeForConditionalGeneration

Holo4 35B A3B — Hardware Requirements & GPU Compatibility

VisionFunctions

H Company's mixture-of-experts vision-language model for computer tasks is Holo4-35B-A3B, built upon Qwen3.6-35B-A3B and featuring about 35 billion total parameters with roughly 3.9 billion active per token. When paired with the hai-agents harness, it interprets screenshots and tool outputs to generate clicks, typed input, code, and API calls for web, desktop, and mobile workflows. Performance on OSWorld 2.0 stands at 30.9% and AutomationBench yields 34.5%—both lower than the dense 27B sibling yet achieved with reduced cost per task. Since only a fraction of weights activate during inference, decoding remains quick; however, all 35 billion parameters still occupy memory space, meaning you'll need either a quantized version for a 24 GB GPU using offloading or a system with plenty of unified memory. The context window supports up to 262,144 tokens. Weights are available under the Apache 2.0 license for commercial deployment, unlike the non-commercial Holo4-27B release. It launched in September 2026, and the model card lists FP8, NVFP4, and Q4 GGUF variants.

35.1B params · MoE3.9B active262K contextApache 2.01.5K downloads38 likes53.4K quant downloads
A

Runs well on a 24 GB GPU

Runs from12 GB IQ2_S · 64KBest on24 GB Q4_K_M · ~171 tok/s · 128KMacs from16 GB IQ2_M · ~57 tok/s

Apache 2.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Holo4 35B A3B

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

Ollama
one command · needs Ollama installed
$ollama run hf.co/bartowski/Hcompany_Holo4-35B-A3B-GGUF:Q4_K_M

Q4_K_M fits fully on the RTX 4090 · 2.5 GB free · 16K context

Download: Hcompany_Holo4-35B-A3B-Q4_K_M.gguf22.32 GBbartowski

install, OS, other quants

Ways to runOther ways

each opens step-by-step

Run in cloud

Your own private copy · Q4_K_M · 24 GB GPU

  • The easiest way to try this model
  • Running in a few minutes, easy setup
  • Chat right here or connect any app

$5+ bonus for new users

How Much VRAM Does Holo4 35B A3B Need?

Select a quantization level; the listed hardware adheres to it.

QuantizationVRAM+32KOn RTX 4090 (reference)
15.3 GB15.6 GB2-bit · K-quant — 2-bit quantization with K-quant improvementsFits: 256K context · ~209 tok/s
15.7 GB16 GB3-bit small — 3-bit small quantizationFits: 256K context · ~205 tok/s
17.5 GB17.8 GB3-bit medium — 3-bit medium quantizationFits: 256K context · ~194 tok/s
17.9 GB18.2 GB4-bit · legacy — 4-bit legacy quantizationFits: 256K context · ~191 tok/s
21.4 GB21.7 GB4-bit medium — 4-bit medium quantization — most popular sweet spotFits: 128K context · ~171 tok/s
25.4 GB25.7 GB5-bit medium — 5-bit medium quantization — good quality/size tradeoffDoes not fit: needs 25.5 GB
29.4 GB29.6 GB6-bit, very good — 6-bit quantization, very good qualityDoes not fit: needs 29.4 GB
35.5 GB35.7 GB8-bit, near-lossless — 8-bit quantization, near-losslessDoes not fit: needs 35.6 GB

Selected Q4_K_M

4-bit medium quantization

21.5GBat 16K context

Weights
21.1 GB
KV cache
0.2 GB
Overhead
0.3 GB

Fits RTX 4090 (reference) · 2.5 GB free · ~171 tok/s

≥50 tok/s needs ≥206 GB/s · RTX 5090 ~236, RTX 3090 Ti ~171

Or .

Which Hardware Can Run Holo4 35B A3B?

Q4_K_M · 21.4 GB

Holo4 35B A3B (Q4_K_M) requires 21.4 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 28+ GB is recommended. Using the full 262K context window can add up to 2.6 GB, bringing total usage to 24.1 GB. 7 GPUs can run it, including NVIDIA GeForce RTX 5090. 112 devices with unified memory can run Holo4 35B A3B, including Mac Studio M5 Ultra (256 GB), Mac Studio M5 Ultra (512 GB), NVIDIA RTX Spark N1X Laptop (32 GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M
8 GB · RTX 4060Doesn't fit — smallest quant needs 10.1 GB8 GB tier →
12 GB · RTX 3060 12GBIQ2_S · 11.4 GB · ~132 tok/s · 64K12 GB tier →
16 GB · RTX 5060 Ti 16GBQ3_K_S · 15.8 GB · ~122 tok/s · 32K16 GB tier →
24 GB · RTX 4090referenceQ4_K_M · 21.5 GB · ~171 tok/s · 128K24 GB tier →
Mac 32 GB · Mac Mini M4Q4_K_M · 21.5 GB · ~28 tok/s · 256KMacs →
Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 21.5 GB · ~70 tok/s · 256KMacs →

Alternatives to Holo4 35B A3B

Similar models

6
ModelScoreVRAMtok/sWhy it is similar
Qwen3.6 35B A3BAlibaba106.822 GB~151same size
MiniCPM V 4.7 35B A3BOpenBMB—22.1 GB~169same size
Ornith 1.5 35B A3Bornith-ai—22 GB~151same size
Agents A1InternScience—21.5 GB~171same size
Intern S2 PreviewInternLM—73.2 GB—needs > 24 GBsame size
EXAONE 4.5 33BLGAI-EXAONE—23.9 GB~33same size

Specifications

Publisher
Hcompany
Parameters
35.1B
Active parameters
3.9B per token
Architecture
Qwen3_5MoeForConditionalGeneration
Context Length
262,144 tokens
Vocabulary Size
248,320
Release Date
2026-09-24
License
Apache 2.0

Frequently Asked Questions

How much VRAM does Holo4 35B A3B need?

Holo4 35B A3B requires 21.4 GB of VRAM at Q4_K_M, or 70.6 GB at BF16. Full 262K context adds up to 2.6 GB (24.1 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 35.1B × 4.8 bits ÷ 8 = 21.1 GB

KV Cache + Overhead ≈ 0.3 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.

KV Cache + Overhead ≈ 2.9 GB (at full 262K context)

VRAM usage by quantization

21.4 GB
24.1 GB

Learn more about VRAM estimation →

Can NVIDIA GeForce RTX 4090 run Holo4 35B A3B?

Yes, at Q4_K_L (21.9 GB) or lower. Higher quantizations like Q5_K_S (24.5 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.

What's the best quantization for Holo4 35B A3B?

For Holo4 35B A3B, Q4_K_M (21.4 GB) offers the best balance of quality and VRAM usage. Q4_K_L (21.9 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 10.0 GB.

VRAM requirement by quantization

IQ2_XXS
10.0 GB
IQ3_XS
14.9 GB
Q4_0
17.9 GB
Q4_K_M ★
21.4 GB
Q4_K_L
21.9 GB
BF16
70.6 GB

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Holo4 35B A3B on a Mac?

Yes — Mac Mini M2 Pro (16 GB) and 145 other Macs can run Holo4 35B A3B, the smallest at IQ2_M. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 21.4 GB of usable unified memory (RAM minus macOS overhead), from Mac Studio M1 Max (32 GB).

Can I run Holo4 35B A3B locally?

Yes — Holo4 35B A3B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Holo4 35B A3B?

At Q4_K_M, Holo4 35B A3B can reach ~118 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~171 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.

tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)

Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.

Example: NVIDIA B200 → 2.4 GB active ÷ (8000 × 0.65) = 0.46 ms, plus 40 layers × 0.055 ms = 2.20 ms, so 1000 ÷ 2.66 ms = ~376 tok/s

Estimated speed at Q4_K_M (21.4 GB)

~376 tok/s
~171 tok/s
~376 tok/s
~339 tok/s

Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.

Learn more about tok/s estimation →

What's the download size of Holo4 35B A3B?

At Q4_K_M, the download is about 21.06 GB. The full-precision BF16 version is 70.21 GB. The smallest option (IQ2_XXS) is 9.65 GB.

Which GPUs can run Holo4 35B A3B?

7 consumer GPUs can run Holo4 35B A3B at Q4_K_M. Top options include AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090.

Which devices can run Holo4 35B A3B?

112 devices with unified memory can run Holo4 35B A3B at Q4_K_M (21.4 GB), including AMD Ryzen 7 8845HS (Hawk Point) Laptop, AMD Ryzen AI 7 350 (Krackan Point) Laptop, AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.