GMKtec EVO-X2 (128GB unified)
The $/GB winner: runs 70B-class and quantized big-MoE (V4 Flash tier) that no consumer GPU fits
H200 cloud pricing, August 2026
Live H200 quotes (141 GB VRAM) across the providers we track, cheapest first. Marketplace prices are real asks from the Vast.ai order book; secure/on-demand tiers carry datacenter SLAs (what the tiers mean).
Quotes refreshed 2026-08-23
Price floor: $2.00/hr (spot) at
Verda.
| Provider | Tier | $/hr per GPU | Median $/hr | Live offers | |
|---|---|---|---|---|---|
| Verda | spot | $2.00 | - | - | |
| Vast.ai | marketplace | $3.29 | $3.95 | 28 | Rent |
| RunPod | community | $3.59 | - | - | Rent |
| Verda | on-demand | $4.00 | - | - | |
| RunPod | secure | $4.59 | - | - | Rent |
Need 8-128 of these? Get a cluster quote →, or
hand us the workload for a ranked deployment plan, free while in beta.
One H200 or two H100 SXMs?
For models needing more than 80 GB but at most 141 GB, a single H200 holds the whole model on one card: no tensor-parallel split, no inter-GPU communication overhead. A pair of H100 SXMs brings more raw compute but pays that overhead. Today's quotes, same provider and tier on both sides (the 2× column doubles the per-GPU price):
| Provider | Tier | 1× H200 $/hr | 2× H100 SXM $/hr | Cheaper |
|---|---|---|---|---|
| Verda | spot | $2.00 | $3.25 | H20038.5% less |
| Vast.ai | marketplace | $3.29 | $2.67 | 2× H100 SXM18.9% less |
| RunPod | community | $3.59 | $5.38 | H20033.3% less |
| Verda | on-demand | $4.00 | $6.50 | H20038.5% less |
| RunPod | secure | $4.59 | $6.58 | H20030.2% less |
From today's quotes: one H200 is cheaper in 4 of 5 comparable provider/tier pairs. Recomputed on every daily refresh.
What fits on a single H200
Models from our tracked open-weight set that fit in 141 GB, with KV-cache headroom included. Bigger models need multi-GPU nodes (math in the calculator).
| Model | Params | Fits at | VRAM needed | Cheapest API $/1M out |
|---|---|---|---|---|
| GLM-4.7 Flash | 31B | FP8 | 43 GB | compare → |
| Qwen3.6 35B A3B | 36B | FP8 | 50 GB | compare → |
| Qwen3.6 27B | 27.8B | FP8 | 39 GB | compare → |
| Qwen3.5 35B A3B | 36B | FP8 | 50 GB | compare → |
| Qwen3.5 27B | 27.8B | FP8 | 39 GB | compare → |
| Qwen3 32B | 32.8B | FP8 | 46 GB | compare → |
| Qwen3 30B A3B | 30.5B | FP8 | 42 GB | compare → |
| Mistral Small 3.1 | 24B | FP8 | 33 GB | |
| Nemotron 3 Nano | 30B | FP8 | 42 GB | compare → |
| IBM Granite 4.1 30B | 30B | FP8 | 42 GB | |
| IBM Granite 4.1 8B | 8.8B | FP8 | 13 GB | compare → |
| GPT-OSS 20B | 21B | FP8 | 29 GB | compare → |
| Llama 3.3 70B | 70B | FP8 | 97 GB | compare → |
| Llama 3.1 8B | 8B | FP8 | 11 GB | compare → |
| Llama 3.2 3B | 3.2B | FP8 | 5 GB | compare → |
| Gemma 3 27B | 27B | FP8 | 38 GB | compare → |
| Gemma 3 12B | 12B | FP8 | 17 GB | compare → |
| Phi-4 | 14.7B | FP8 | 21 GB | compare → |
| Mixtral 8x7B | 46.7B | FP8 | 65 GB | compare → |
| DeepSeek R1 Distill Llama 70B | 70B | FP8 | 97 GB | compare → |
| DeepSeek R1 Distill Qwen 32B | 32.8B | FP8 | 46 GB | compare → |
| GLM-4.5 Air | 106B | INT4 | 73 GB | compare → |
| Mistral Medium 3.5 | 128B | INT4 | 88 GB | compare → |
| Mistral Small 4 | 119.4B | INT4 | 83 GB | |
| Leanstral 1.5 | 119B | INT4 | 82 GB | |
| Nemotron 3 Super | 120B | INT4 | 83 GB | compare → |
| GPT-OSS 120B | 117B | INT4 | 81 GB | compare → |
| Llama 4 Scout | 109B | INT4 | 75 GB | compare → |
Can't buy a H200? The desk-side alternatives
H200 cards only ship in servers, but the "own your inference box" itch has real answers: run the economics in the calculator, then look at what this audience actually buys:
NVIDIA RTX 5090 (32GB)
Bandwidth-per-dollar king for the A3B-MoE and 27-32B dense workhorses
NVIDIA RTX PRO 6000 Blackwell (96GB)
The startup standard: only sub-$10K single card that runs 70B+ comfortably
As an Amazon Associate we earn from qualifying purchases · hardware links may be referral links · picks are editorial, not paid · disclosure
FAQ
How much does it cost to rent a H200?
As of 2026-08-23, H200 rentals start at $2.00/hr (spot tier at Verda). Datacenter capacity with SLAs typically costs more than community or marketplace hardware.
What can you run on a H200?
With 141 GB of VRAM, a single H200 fits models up to roughly 100B parameters at FP8 or 201B at INT4 quantization, including KV-cache headroom.
Cite or embed today's floor
Live badge for a README. It updates itself with every daily refresh:
[](https://llmhosting.ai/gpus/h200)
Citation: H200 rental floor $2.00/hr (llmhosting.ai, 2026-08-23). Free to reuse with a link back.