Hy4 has 774.4B parameters (48B active per token). In BF16 its weights alone take 1436 GiB; quantized to 4-bit, about 433 GiB. Each 32K-token request adds 2.83 GiB of KV cache. Cheapest way to run it today: 7 × NVIDIA A100 80GB at $5.91 per hour. It reads text, images and video, calls tools and lets you set its reasoning effort.
Weights plus the KV cache of one request (BF16 cache), in GiB. Add 1 to 3 GiB for the inference engine itself.
Weights format
Weights only
+ 8K context
+ 32K context
+ 128K context
+ 1M context
BF16 / FP16
1436
1437
1439
1448
1527
FP8 / INT8
720
720
722
731
810
4-bit (GGUF Q4_K_M)
433
434
436
444
524
Hy4 GGUF files
Exact size of each quantization in Xylolabs/Hy4-GGUF (opens in a new tab), and the smallest GPU that runs it with llama.cpp, Ollama or LM Studio and an 8K-token context.
Quantization
File size
Smallest GPU
Q8_0
762 GiB
More than 96 GB: see the calculator
Q4_K_M
432 GiB
More than 96 GB: see the calculator
IQ3_S
306 GiB
More than 96 GB: see the calculator
IQ2_XXS
204 GiB
More than 96 GB: see the calculator
Read every day from Hugging Face. Split files are added together; vision projectors are left out.
How many GPUs to run Hy4
Fewest GPUs that fit the weights plus one 8K-token request. Data-center GPUs use vLLM (90% of memory usable), the others llama.cpp. Several GPUs split the model with tensor or pipeline parallelism.
GPU
BF16 / FP16
8-bit
4-bit
GeForce RTX 4090 24 GB
96 (several machines)
48 (several machines)
24 (several machines)
GeForce RTX 5090 32 GB
64 (several machines)
32 (several machines)
16 (several machines)
RTX PRO 6000 Blackwell 96 GB
16 (several machines)
8
5
Mac M5 Max 128 GB 96 GB usable
16 (several machines)
8
5
Ryzen AI Max+ 395 128 GB 96 GB usable
16 (several machines)
8
5
DGX Spark (GB10) 128 GB
16 (several machines)
6
4
NVIDIA H100 80 GB
24 (several servers)
16 (several servers)
7
NVIDIA H200 141 GB
16 (several servers)
6
4
NVIDIA B200 (HGX) 180 GB
16 (several servers)
5
3
8-bit: FP8 with vLLM (INT8 on GPUs without FP8), Q8_0 with llama.cpp. 4-bit: AWQ / GPTQ with vLLM, Q4_K_M with llama.cpp, both counted at the Q4_K_M size.
Cheapest way to run Hy4 today
For each GPU: the fewest cards that fit Hy4 with one 8K-token request, the most faithful weight format at that count, and the cheapest on-demand price (2026-11-22).
On-demand prices from AWS, Google Cloud, Azure, RunPod, Vast.ai, Oracle Cloud, OVHcloud, Scaleway, Verda, Akash and UpCloud, checked every hour. Serving many users needs more KV cache, so more memory: size it in the calculator. Some links are affiliate links: StudioTV may earn a commission, at no extra cost to you.
vLLM and SGLang apply these defaults unless a request sets its own. With llama.cpp, pass them yourself: --temp 0.9 --top-p 1
Reasoning. Adjustable reasoning effort. Reasoning tokens take context and KV cache like any other token: leave room for them.
KV cache: how memory grows with context
Architecture: 78 layers · MLA (512 + 64 latent) on 78 · indexer keys on 21. Its cache stays compact because MLA stores one compressed latent per token and layer instead of full keys and values. Only 48B parameters are active per token, but all 774B must sit in memory: VRAM depends on the total, speed on the active part.
Context per request
BF16 cache
FP8 cache
8K tokens
0.71 GiB
0.41 GiB
32K tokens
2.83 GiB
1.65 GiB
128K tokens
11.3 GiB
6.58 GiB
1M tokens
90.5 GiB
52.7 GiB
Hy4 VRAM FAQ
How much VRAM does Hy4 need?
In BF16 the weights alone take 1436 GiB (1542 GB). In FP8 that is 720 GiB, and about 433 GiB with 4-bit quantization (Q4_K_M). Each request then adds KV cache: 0.71 GiB at 8K tokens and 2.83 GiB at 32K (BF16 cache).
Can Hy4 run on a single RTX 4090 (24 GB)?
No. Even in 4-bit it needs 24 RTX 4090s, spread over several machines. In 4-bit it needs 7 H100 GPUs.
Can Hy4 run on an RTX 5090 (32 GB)?
No. Even in 4-bit it needs 16 RTX 5090s, spread over several machines.
Can Hy4 run on a Mac?
No single Mac holds it, even in 4-bit: it needs 434 GiB with an 8K-token context, and the largest Mac (512 GB) gives its GPU about 384 GB.
Can Hy4 run on a DGX Spark or a Ryzen AI Max+ 395 (128 GB)?
No. Even in 4-bit it needs 4 DGX Spark units or 5 Ryzen AI Max+ 395 machines (128 GB), linked over the network.
What is the cheapest way to run Hy4?
On 2026-11-22, the cheapest on-demand setup is 7 × NVIDIA A100 80GB with 4-bit weights on Vast.ai, at $5.91 per hour (about $4,312.84 per month). Next: 5 × RTX PRO 6000 Blackwell (96 GB) with Q4_K_M weights on Vast.ai, at $6.68 per hour. Sized for one 8K-token request; prices are checked every hour.
How many H100 GPUs does Hy4 need?
With one request and an 8K-token context: 24 in BF16, 16 in FP8 and 7 in 4-bit (vLLM, 90% of memory usable). Above 8 GPUs, that means several servers. Serving many users at once needs more memory for their KV caches: the calculator sizes that for you.
Can Hy4 read images?
Yes. Hy4 reads text, images and video. Images add tokens to the context, so they also add KV cache.
Does Hy4 support tool calling?
Yes, Hy4 supports tool (function) calling: it can ask your application to run a function and use the result.
Does Hy4 have a thinking mode?
Adjustable reasoning effort. Reasoning tokens count toward the context and the KV cache.
What temperature should I use with Hy4?
The official defaults are temperature 0.9, top-p 1.
Hy4 VRAM badge
For a model card or a README: the 4-bit size and the smallest GPU it fits on, linked to this page. Data as JSON: /api/models/hy4.json.
Estimates, not guarantees: computed from the official config.json with the same engine as the LLM VRAM Calculator. Real usage depends on your engine version and settings.