Qwen3-VL-8B (FP8): fast vision on a DGX Spark
Qwen3-VL-8B-FP8, the vision model, filling the fast auxiliary roles on the DGX Spark (GB10) — titles, approvals, MCP routing, screenshot follow-ups.
| Metric | Value |
|---|---|
| Status | ✅ up, no errors |
| Generation throughput | 11–25 tok/s (peaks 25) |
| Prompt throughput | ~76 tok/s sustained |
| KV cache usage | 2–3% (massive headroom) |
| MM (image) cache hit rate | 60–75% |
| Prefix cache hit rate | 12–60% |
| Warnings | 2× Triton JIT first-touch (one-time, harmless) |
Qwen is actually faster at generation than the 120B Nemotron (25 vs ~15 tok/s) — exactly what you want where latency matters. The multimodal cache working well means repeated questions about the same image get a free speedup: the vision encoder skips the re-encode.
One thing to watch: system memory was tight (≈1.9 GiB free). Both engines are stable but with little slack. If real OOMs ever show up, drop this one to a smaller GPU-memory fraction and a 16k context to free ~5 GiB.