Skip to content

Muse Spark Open 816B A42B on 4× H200 NVL: VRAM, speed and fit

Muse Spark Open 816B A42B at 4-bit runs on the 4× H200 NVL with 1M of context at about 89 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 4× H200 NVL · 4-bit (Q4_K_M) · Ollama / LM Studio · Chat

GPU memory
532 GB
of 564 GB
Context
1M
auto
Speed
89 tok/s
31 when full
First reply
0.7 s
2K prompt

Where the memory goes

Model
464 GB
Context
64.4 GB
Overhead
3.4 GB
OS
0.0 GB
Free
32.1 GB
Memory
532 GB of 564 GB
Context
1M tokens · auto
Speed
89 tok/s · 31 when full
First reply
0.7 s for a 2K prompt
Running cost
$33/mo at home · $1,162/mo on a cloud 4× MI300X

Validation: Validated

What Muse Spark Open 816B A42B on 4× H200 NVL was checked against
CheckTargetResultStatus
Fits in GPU memory564 GB usable532 GB neededPass
Headroom5% free32.1 GB free (6%)Pass
Context for chat16K or more1MPass
Writing speed15 tok/s or more89 tok/sPass
First replyunder 3 s, 2K prompt0.7 sPass

Targets: a reply you can read along with, a first reply the use case tolerates, and 5% of memory left free. Calculated from published specs, not benchmarked.

The setup

Model
Muse Spark Open 816B A42B
Parameters
822B, 42B active
Precision
4-bit (Q4_K_M)
Context
1M, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
4× H200 NVL
GPU can use
564 GB of 564 GB

Running cost

At home
$33/mo
1520 W at the wall
Renting instead
$1,162/mo
4× MI300X at $9.56/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 21, 2026.

How to run it

llama.cpp
llama-server -hf unsloth/Muse-Spark-Open-816B-A42B-GGUF
vLLM
vllm serve meta-models/Muse-Spark-Open-816B-A42B
MLX (Mac)
mlx_lm.chat --model mlx-community/Muse-Spark-Open-816B-A42B-4bit

Muse Spark Open 816B A42B on similar machines

Other models that run on 4× H200 NVL

Questions

Can Muse Spark Open 816B A42B run on 4× H200 NVL?

Muse Spark Open 816B A42B at 4-bit runs on the 4× H200 NVL with 1M of context at about 89 tokens a second.

Report VR-16BP596, calculated Nov 22, 2026 from published specs, not benchmarked.