llmrun
TencentHYV4ForCausalLM

Hy4 Preview — Hardware Requirements & GPU Compatibility

Chat

Hy4 Preview is a 780.0B-parameter open language model from Tencent. It supports a context window of up to 1,048,576 tokens. At BF16 it needs about 1560.71 GB of VRAM — see which GPUs and Macs can run it below.

780.0B params · MoE50B active1049K contextApache 2.016.4K downloads483 likes

Hy4 Preview runs on a GPU cluster

BF16 needs ~1.6 TB · about 3 nodes of 8× H100 SXM 80 GB

Apache 2.0

Not set up for it?

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Hy4 Preview

One recommended way. Every other method, OS and quant is in the step-by-step guide.

Recommended

vLLM on GPU nodes
tensor + pipeline parallel · needs vLLM on every node
$vllm serve tencent/Hy4-preview --max-model-len 16384 --tensor-parallel-size 8 --pipeline-parallel-size <nodes>

Size the nodes from the table above

Download: Weights on Hugging Face

Ray cluster setup, other ways

Other ways

each opens step-by-step

Run in cloud

Rent 3× 8-GPU H100 SXM nodes · BF16

Estimated hourly price
~$53.07/hr
est., whole cluster
GPUs
24 GPUs
1.9 TB of VRAM
Deployon Vast.aiRent H100 SXM$2.21/hr per GPU (opens in new tab)
Compare GPUs →

How Much VRAM Does Hy4 Preview Need?

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

QuantizationVRAM+32KRuns on
1560.7 GB1568.1 GB16-bit · bfloat — Brain floating point 16 — preferred for trainingRuns on 3 nodes of 8× H100 SXM
est.calculated, no published file yet

Selected BF16

Brain floating point 16

1564.2GBat 16K context

Weights
1559.9 GB
KV cache
3.9 GB
Overhead
0.3 GB

Server-class: ~1.6 TB at 16K context · about 3 nodes of 8× H100 SXM 80 GB

Which Hardware Can Run Hy4 Preview?

BF16 · 1560.7 GB

Hy4 Preview (BF16) requires 1560.7 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 2029+ GB is recommended. Using the full 1049K context window can add up to 250.8 GB, bringing total usage to 1811.5 GB. No consumer GPU has enough memory.

Needs server-class hardware

Hy4 Preview runs on a GPU cluster · BF16 needs ~1.6 TB

Hardware that runs this model
HardwareQuantMemoryContextHardware page
GPU cluster · 3 nodes of 8× H100 SXM 80 GBBF161564.2 GB64K ctxH100 SXM →BF16 · 1564.2 GB · 64K ctx

≤ 640 GB tier →

By VRAM tier · typical GPUs and Macs at BF16

Consumer GPUs (up to 32 GB) and Macs up to 512 GB: too small for this model (needs ≥ 1.6 TB)8 GB tier →12 GB tier →16 GB tier →24 GB tier →Macs →

How good is Hy4 Preview?

llmrun Score blends public benchmark boards into one number; 100 = the average open model.

Not enough data yet for an llmrun Score.

A Score needs results on enough independent boards. How the Composite Score is built

Benchmarks

Here are all 2 boards that have results for this model, sorted by their weight in the llmrun Score; badges show who ran the test.

  • LMArena WebDev

    LMArena · Oct 2026Arena

    1631.5

    #3/52 open

    Arena
  • ProofBench

    Epoch AI · Aug 2026Lab-reported

    75.0%

    #2/26 open

    Lab-reported
All 2 Hy4 Preview benchmarks

Alternatives to Hy4 Preview

Similar models

3
ModelScoreVRAMWhy it is similar
GLM 5.3Z.ai130.1453.8 GBsame size
DeepSeek V4.1 FlashDeepSeek129.5458.4 GBsame size
Muse Spark Open 816B A42BMeta125.3551.9 GBsame size

Specifications

Publisher
Tencent
Parameters
780.0B
Active parameters
49.7B per token
Architecture
HYV4ForCausalLM
Context Length
1,048,576 tokens
Vocabulary Size
120,832
Release Date
2026-08-27
License
Apache 2.0

Resources

Frequently Asked Questions

How much VRAM does Hy4 Preview need?

Hy4 Preview requires 1560.7 GB of VRAM at BF16. Full 1049K context adds up to 250.8 GB (1811.5 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 780.0B × 16 bits ÷ 8 = 1559.9 GB

KV Cache + Overhead ≈ 0.8 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 ≈ 251.6 GB (at full 1049K context)

VRAM usage by quantization

1560.7 GB
1811.5 GB

Learn more about VRAM estimation →

Can NVIDIA GeForce RTX 5090 run Hy4 Preview?

No — Hy4 Preview requires at least 1560.7 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.

Can I run Hy4 Preview on a Mac?

Hy4 Preview requires at least 1560.7 GB at BF16. None of the 146 Macs we list has that much usable memory (the largest has 508.5 GB after macOS overhead).

Can I run Hy4 Preview locally?

Yes — Hy4 Preview can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

What's the download size of Hy4 Preview?

At BF16, the download is about 1559.92 GB.

Which GPUs can run Hy4 Preview?

No single consumer GPU has enough VRAM to run Hy4 Preview at BF16 (1560.7 GB). Multi-GPU or professional hardware is required.

Which devices can run Hy4 Preview?

Hy4 Preview requires at least 1560.7 GB at BF16, which exceeds the usable memory of every unified-memory device we list (the largest has 508.5 GB).