Hy3 REAM 100B — Hardware Requirements & GPU Compatibility
Chat
Hy3 REAM 100B is a 104.1B-parameter open language model from Akicou in the Hunyuan 3 family. It supports a context window of up to 262,144 tokens. At Q4_K_M it needs about 63.12 GB of VRAM — see which GPUs and Macs can run it below.
Hy3 REAM 100B runs on a 96 GB+ Mac or a workstation GPU
Best local pick → Mac Studio (M3 Ultra, 96GB) · Q4_K_M · 65.5 GB · ~27 tok/s
Macs from 64 GB (Q3_K_M · ~30 tok/s)
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Hy3 REAM 100B
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve Akicou/Hy3-REAM-100B --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Fits on A100 80GB (14 GB headroom) · Q4_K_M
- Generation speed
- ~75 tok/s
- generation speed
- Cost per 1M output tokens
- $4.93
- per 1M output tokens
How Much VRAM Does Hy3 REAM 100B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 44.9 GB | 49.9 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Does not fit: needs 47.2 GB | |
| 51.4 GB | 56.4 GB | 3-bit medium — 3-bit medium quantization | Does not fit: needs 53.8 GB | |
| 63.1 GB | 68.2 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Does not fit: needs 65.5 GB | |
| 74.8 GB | 79.9 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Does not fit: needs 77.2 GB | |
| 86.5 GB | 91.6 GB | 6-bit, very good — 6-bit quantization, very good quality | Does not fit: needs 88.9 GB | |
| 104.8 GB | 109.8 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Does not fit: needs 107.1 GB | |
| 208.9 GB | 213.9 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 211.3 GB |
Selected Q4_K_M
4-bit medium quantization65.5GBat 16K context
- Weights
- 62.5 GB
- KV cache
- 2.7 GB
- Overhead
- 0.3 GB
Doesn't fit RTX 4090 (reference) · needs 41.5 GB more
Which Hardware Can Run Hy3 REAM 100B?
Q4_K_M · 63.1 GBHy3 REAM 100B (Q4_K_M) requires 63.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 83+ GB is recommended. Using the full 262K context window can add up to 42.6 GB, bringing total usage to 105.7 GB. No consumer GPU has enough memory. 31 devices with unified memory can run Hy3 REAM 100B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Studio (M3 Ultra, 96GB).
64–80 GB
tight · up to 4K ctx
1 device family
Orin~11Simulate output on NVIDIA Jetson AGX Orin 64GB96 GB
fits · 30.5 GB free · up to 128K ctx
2 Mac classes
M3 Ultra*96 GB~27Simulate output on Mac Studio (M3 Ultra, 96GB)M3 Max*96 GB~14Simulate output on MacBook Pro 14" M3 Max (96 GB)
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Llama 4 Scout 17B 16E InstructMeta | 81.8 | 77 GB | —needs > 24 GB | same size |
| C4ai Command A 03 2025Cohere | 79.8 | 78.7 GB | —needs > 24 GB | same size |
| GLM 4.6VZ.ai | — | 66 GB | —needs > 24 GB | same size |
| Ring Flash 2.0Inclusion AI | — | 63.1 GB | —needs > 24 GB | same size |
| Ling Flash 2.0Inclusion AI | — | 63.1 GB | —needs > 24 GB | same size |
| C4ai Command R Plus 08 2024Cohere | — | 73.8 GB | —needs > 24 GB | same size |
Specifications
- Publisher
- Akicou
- Family
- Hunyuan 3
- Parameters
- 104.1B
- Active parameters
- 19.6B per token
- Architecture
- HYV3ForCausalLM
- Context Length
- 262,144 tokens
- Vocabulary Size
- 120,832
- Release Date
- 2026-07-13
Frequently Asked Questions
How much VRAM does Hy3 REAM 100B need?
Hy3 REAM 100B requires 63.1 GB of VRAM at Q4_K_M, or 208.9 GB at BF16. Full 262K context adds up to 42.6 GB (105.7 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 104.1B × 4.8 bits ÷ 8 = 62.5 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.
KV Cache + Overhead ≈ 43.2 GB (at full 262K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run Hy3 REAM 100B?
No — Hy3 REAM 100B requires at least 44.9 GB at Q2_K, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for Hy3 REAM 100B?
For Hy3 REAM 100B, Q4_K_M (63.1 GB) offers the best balance of quality and VRAM usage. Q5_K_M (74.8 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 44.9 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Hy3 REAM 100B on a Mac?
Yes — Mac Pro M2 Ultra (64 GB) and 41 other Macs can run Hy3 REAM 100B, the smallest at Q3_K_M. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 63.1 GB of usable unified memory (RAM minus macOS overhead), from Mac Studio (M3 Ultra, 96GB).
Can I run Hy3 REAM 100B locally?
Yes — Hy3 REAM 100B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Hy3 REAM 100B?
At Q4_K_M, Hy3 REAM 100B can reach ~54 tok/s on AMD Instinct MI350X. 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 → 11.9 GB active ÷ (8000 × 0.65) = 2.28 ms, plus 80 layers × 0.055 ms = 4.40 ms, so 1000 ÷ 6.68 ms = ~150 tok/s
Estimated speed at Q4_K_M (63.1 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Hy3 REAM 100B?
At Q4_K_M, the download is about 62.48 GB. The full-precision BF16 version is 208.27 GB. The smallest option (Q2_K) is 44.26 GB.
Which GPUs can run Hy3 REAM 100B?
No single consumer GPU has enough VRAM to run Hy3 REAM 100B at Q4_K_M (63.1 GB). Multi-GPU or professional hardware is required.
Which devices can run Hy3 REAM 100B?
31 devices with unified memory can run Hy3 REAM 100B at Q4_K_M (63.1 GB), including ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB), GMKtec EVO-X2 (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.