H100 80GB vs H200 141GB
The H200 141GB holds 8 models from our catalog that the H100 80GB cannot fit at Q4_K_M, and decodes about 1.4× faster on memory bandwidth. Decode speed on a memory-bound workload scales with bandwidth, so that ratio is the number that matters once the model fits at all.
Specs side by side
| Spec | H100 80GB | H200 141GB |
|---|---|---|
| VRAM | 80 GB | 141 GB |
| Memory bandwidth | 3350 GB/s | 4800 GB/s |
| TDP | 700 W | 700 W |
| Models that fit (Q4_K_M) | 473 of 506 | 481 of 506 |
Form factor: H100 80GB figures are for the SXM5 module, H200 141GB figures are for the SXM module. PCIe/NVL variants differ in bandwidth and TDP — see the H100 80GB page and methodology.
What the extra VRAM actually buys
8 models in our catalog fit on the H200 141GB at Q4_K_M but not on the H100 80GB. The largest of them:
Falcon 180BFalcon 180B ChatBLOOM 176BBLOOM 176B zMixtral 8x22BMixtral 8x22B Instruct
Largest model each can run
At Q4_K_M the biggest fit in our catalog is Pixtral Large 124B Instruct (74.3 GB) on the H100 80GB, and Falcon 180B (107.5 GB) on the H200 141GB. Both figures are weights plus runtime overhead; KV cache for your context is extra.
Power and running cost
At 8 hours a day the H100 80GB draws about 168 kWh a month against 168 kWh for the H200 141GB — a difference of 0 kWh. Multiply by your own tariff; we do not assume one. Board TDP is the rated figure, not measured draw.
FAQ
H100 80GB vs H200 141GB: which should I buy for local AI?
The H200 141GB fits 481 of the 506 models in our catalog at Q4_K_M against 473 for the H100 80GB, and has about 1.4× the memory bandwidth, which is what sets decode speed once a model fits. If the models you run already fit the H100 80GB, the extra spend buys context length and speed, not a new model tier.
Is the H200 141GB faster than the H100 80GB?
On memory-bound decoding, yes — roughly 1.4× the throughput, since token generation scales with memory bandwidth (4800 GB/s vs 3350 GB/s). This is an upper bound from bandwidth alone, not a measured benchmark.
How much more VRAM does the H200 141GB have?
61 GB more (141 GB vs 80 GB). That difference lets it hold 8 additional models from our catalog at Q4_K_M.
Need a third card in the mix? The interactive compare tool lines up any three. To size KV cache for your context length, use the GPU Memory Calculator.
Related matchups
Other head-to-heads involving these cards:
A100 80GB vs H100 80GBRTX 4090 vs H100 80GBRTX 5090 vs H100 80GBA100 80GB vs H200 141GB
New to sizing? What VRAM is and why bandwidth sets decode speed cover the basics, and quantization explains the formats in the table above. Common mistakes when sizing VRAM lists the traps.