microsoft/Phi-3-medium-4k-instruct
14B params · est. 8.4 GB Q4
NVIDIA H200 141GB: likely fitNVIDIA H100 SXM5 80GB: likely fit
Head-to-head
NVIDIA H200 141GB vs NVIDIA H100 SXM5 80GB
Dedicated comparison page for two real hardware profiles. This is built from your benchmark database, trust metadata, and buyer flow instead of generic spec-sheet comparisons.
Device profile
NVIDIA H200 141GB
405B-class models at FP8 with context headroom — the capacity play for frontier open-weight models
Curated Aggregate·2026-04-20141GB / $30k
Best LLM
540 tok/s
VRAM
141 GB
TDP
700 W
Rating
9.4
Device profile
NVIDIA H100 SXM5 80GB
70B FP16 with substantial context, 405B Q4 with headroom. Production multi-tenant serving at batch 8-64 via TensorRT-LLM or vLLM.
Curated Aggregate·2026-04-0980GB / $25k
Best LLM
612 tok/s
VRAM
80 GB
TDP
700 W
Rating
9.4
Quick verdict
NVIDIA H200 141GB scores 9.4 while NVIDIA H100 SXM5 80GB scores 9.4 on MyAIHardware's composite rating.
NVIDIA H200 141GB wins VRAM capacity.
Power draw is effectively tied on paper.
On shared workload evidence, Llama 3 8B Q4 is benchmarked at 215 tok/s for NVIDIA H200 141GB and 178 tok/s for NVIDIA H100 SXM5 80GB.
| Metric | NVIDIA H200 141GB | NVIDIA H100 SXM5 80GB |
|---|---|---|
| MyAI rating | 9.4 | 9.4 |
| Best LLM | 540 tok/s | 612 tok/s |
| VRAM | 141 GB | 80 GB |
| TDP | 700 W | 700 W |
| MSRP | $30k | $25k |
| Workloads | 11 | 13 |
Shared benchmark rows
Llama 3 8B Q4
Llama 3 8B Q4
NVIDIA H200 141GB
215 tok/s
Q4_K_M / 141GB / 2025-02-04
NVIDIA H100 SXM5 80GB
178 tok/s
Q4_K_M / 80GB / 2024-11-08
Mistral 7B Q4
Mistral 7B
NVIDIA H200 141GB
305 tok/s
Q4_K_M / 141GB / 2026-01-28
NVIDIA H100 SXM5 80GB
295 tok/s
Q4_K_M / 80GB / 2025-05-08
Gemma 2 9B Q4
Gemma 2 9B
NVIDIA H200 141GB
220 tok/s
Q4_K_M / 141GB / 2024-07-30
NVIDIA H100 SXM5 80GB
215 tok/s
Q4_K_M / 80GB / 2024-10-14
DeepSeek-R1 Distill 7B
DeepSeek-R1 7B
NVIDIA H200 141GB
305 tok/s
Q4_K_M / 141GB / 2026-04-20
NVIDIA H100 SXM5 80GB
245 tok/s
Q4_K_M / 80GB / 2024-08-04
Qwen 2.5 14B Q4
Qwen 2.5 14B
NVIDIA H200 141GB
178 tok/s
Q4_K_M / 141GB / 2024-12-15
NVIDIA H100 SXM5 80GB
168 tok/s
Q4_K_M / 80GB / 2024-10-21
Llama 3 70B Q4
Llama 3 70B Q4
NVIDIA H200 141GB
84.0 tok/s
Q4_K_M / 141GB / 2024-04-06
NVIDIA H100 SXM5 80GB
66.0 tok/s
Q4_K_M / 80GB / 2024-06-24
External quality layer
Best open models likely to fit this class
This shortlist cross-references the external OpenEvals snapshot with an approximate Q4-class VRAM estimate. Use it to sanity-check whether the hardware you are comparing can host strong open models, not just benchmark toy workloads.
91.0
Qwen/Qwen2-72B
73B params · est. 43.6 GB Q4
NVIDIA H200 141GB: likely fitNVIDIA H100 SXM5 80GB: likely fit
microsoft/Phi-3.5-mini-instruct
3.8B params · est. 2.3 GB Q4
NVIDIA H200 141GB: likely fitNVIDIA H100 SXM5 80GB: likely fit
internlm/internlm2_5-7b-chat
7.7B params · est. 4.6 GB Q4
NVIDIA H200 141GB: likely fitNVIDIA H100 SXM5 80GB: likely fit