OpenModelsHQ

Hy3-preview

Tencent HunyuanLLMMoECustom licenseChina Desk

Hy3-preview is an open-weight language model from Tencent Hunyuan with 299B parameters (mixture of experts). At 4-bit it needs about 200.2 GB of memory, so the smallest device in our directory that fits it is the Rubin, which is announced but not shipping yet.

Overview

Estimated memory at 4-bit
200.2GB

Includes room for an 8k context and runtime overhead. More context needs more memory.

Total parameters
299B
Mixture of experts, all weights counted
Context window
256K
262,144 tokens
Cheapest API, blended / 1M
$0.285
₹27 · GMICloud · 3:1 input to output
All Tencent Hunyuan API prices

Key information

Lab
Tencent Hunyuan
Released
13 Apr 2026
Parameters
299B · MoE
Type
Language
Context
262,144 tokens
HF downloads, 30 days
76,696
View all

Will it run on your hardware?

How we estimate
Memory needed vs devices
Hy3-preview needs about 200.2 GB at 4-bit, 351.3 GB at 8-bit and 660.3 GB at 16-bit. At 4-bit it fits none of the reference devices. Dashed lines are usable memory on each device; a model fits when its estimate stays within 90% of it. Log scale. 8k context.
Memory needed vs devices: data
Estimate or deviceMemory
Q4200.2 GB
Q8351.3 GB
BF16660.3 GB
RTX 5060 Ti 16 GB16 GB usable
RTX 5090 32 GB32 GB usable
DGX Spark 128 GB108.8 GB usable
Mac Studio M5 Ultra 256 GB192 GB usable

Single-device memory estimate. Pick a device to update the On your device column.

36 GB usable memory

FormatMemory neededOn your deviceExample hardware
4-bit (Q4)≈ 200.2 GBToo big 556%RubinAnnounced288 GB memory
8-bit (Q8)≈ 351.3 GBToo big 976%No single device in our directory has enough memory
16-bit (BF16)≈ 660.3 GBToo big 1834%No single device in our directory has enough memory

Fits means the estimate stays within 90% of usable memory. Tight is above 90% and up to 100%. Too big is above 100%. These are memory checks, not speed or compatibility tests.

Mac allowances are 66% of total RAM up to 36 GB, and 75% above. AI boxes allow 85% because their memory is shared with the operating system.

Where to run it

On a Mac

Start with the MLX files, built for Apple Silicon.

mlx-community/Hy3-preview-4bitCheck your Mac

With GGUF

No trusted GGUF build found yet.

Find compatible hardware

Through an API

GMICloud (bf16) is the cheapest provider we found.

$0.18 / $0.60

Input / output per 1M tokens
₹17 / ₹57 at ₹94.29 per $1

Blended: $0.285 (₹27) per 1M. Ranked by a 3:1 mix of input and output tokens.

MLX on Apple SiliconView repository
mlx_lm.generate --model mlx-community/Hy3-preview-4bit \
  --prompt "Explain how a transformer works."

Example commands built from the repository ids above. They need a working install of MLX. We have not run them; check each repository for available files.

Hosted API 1

USD and INR per million tokensAll Tencent Hunyuan API prices
Precision
GMICloudCheapest$0.285₹27$0.18₹17$0.60₹57262,144BF16
View on OpenRouterEndpoint sourceSorted by the blended price: 3 parts input to 1 part output. Context limits differ by provider.

INR at ₹94.29 per $1. Prices fetched 14 Nov 2026.

More about this model

Frequently asked questions

How much memory does Hy3-preview need? At an 8k context, about 200.2 GB at 4-bit, 351.3 GB at 8-bit and 660.3 GB at 16-bit (BF16). A longer context needs more.

Can I run Hy3-preview on a Mac? No current Mac fits the 4-bit estimate of 200.2 GB. The smallest Mac in our directory that does is the Mac Studio M3 Ultra (512GB), which is discontinued. An MLX build is available: mlx-community/Hy3-preview-4bit.

Does the full context fit? Hy3-preview supports up to 262,144 tokens of context. Our memory estimates use 8k, so long prompts need more memory than shown.

What does it cost through an API? The cheapest provider we found is GMICloud, at $0.18 per 1M input tokens and $0.60 per 1M output tokens (₹17 and ₹57 at 94.29 rupees per dollar). Ranked by a 3:1 mix of input and output tokens.

License in plain English

Custom license. This model uses a custom license rather than a standard one. Read it on the model card before using the model commercially. This summary is not legal advice.

Read the license on the model card

OMHQ RigBench

Not yet tested. We show no speed or quality score without a measurement.

Revision history and sources

Last updated . Sources:Hugging Face model card, MLX repository, OpenRouter.