Can SmolVLM2 500M Video Instruct run on the H200?
Yes. At 8K context the best fit is F16, using about 1.6 GB of the 134 GB usable on the H200. Estimated speed ceiling: ~5851 tokens/sec.
+ 0.1 GB with image input. The figures on this page are for text only. With image input enabled, the vision projector (mmproj) file loads alongside the model: F16 at 8K context then needs about 1.7 GB (still fits). llama.cpp's -hf downloads and loads it automatically unless you add --no-mmproj. Calculator with image input on.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | Max tok/s (est., 8K) |
|---|---|---|---|
| 2-bit (Q2_K) | 0.8 GB Fits | 1.0 GB Fits | 25223 |
| 3-bit (Q3_K_M) | 0.9 GB Fits | 1.0 GB Fits | 19352 |
| 4-bit (Q4_K_M) | 0.9 GB Fits | 1.1 GB Fits | 15474 |
| 5-bit (Q5_K_M) | 1.0 GB Fits | 1.1 GB Fits | 13275 |
| 6-bit (Q6_K) | 1.0 GB Fits | 1.2 GB Fits | 11535 |
| 8-bit (Q8_0) • | 1.1 GB Fits | 1.2 GB Fits | 10989 |
| 16-bit (F16/BF16) • | 1.4 GB Fits | 1.6 GB Fits | 5851 |
• real GGUF file size from ggml-org/SmolVLM2-500M-Video-Instruct-GGUF. Speed is a bandwidth ceiling (4800 GB/s), not a benchmark. Methodology.
What fits instead
- Cheapest listed GPU that runs SmolVLM2 500M Video Instruct: Arc B580 (12 GB, launch MSRP $249) at F16, fully in VRAM at 8K context.
- Largest popular model that fits the H200: Qwen3.8 Flash Next (180B, MoE) at Q4_K_M, about 112 GB at 8K context.
"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.
More
SmolVLM2 500M Video Instruct on every GPU and Mac · Every model that runs on the H200 · Other models that fit 64 GB VRAM · Try other settings in the calculator
Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs
Data: Hugging Face Hub (HuggingFaceTB/SmolVLM2-500M-Video-Instruct), NVIDIA specs. Model data last checked 2026-11-18; refreshed hourly.