Can you run Qwen3.5 9B Instruct (Q8) on RTX 5090?
Qwen3.5 9B Instruct (Q8) Q8_0 on the NVIDIA GeForce RTX 5090: verdict, VRAM math and estimated speed.
Yes, the NVIDIA GeForce RTX 5090 runs Qwen3.5 9B Instruct (Q8). 10.7 GB weights at Q8_0 vs 28.8 GB usable VRAM; ~81 tok/s est. (ModelFit, 2026).
VRAM math and speed are ModelFit engine estimates, not measurements. Commands are registry-verified Ollama tags.
Cite this page: ModelFit, Qwen3.5 9B Instruct (Q8) on RTX 5090, https://modelfit.io/can-i-run/qwen3.5-9b-q8-on-rtx-5090/, updated September 2026, CC BY 4.0.
Last updated: September 24, 2026 · Editor: ModelFit Team
What limits this combo
The four constraints the fit engine checks for Qwen3.5 9B Instruct (Q8) on the RTX 5090, in order of what usually breaks first.
Where the RTX 5090 sits for Qwen3.5 9B Instruct (Q8)
On the RTX 5090, Qwen3.5 9B Instruct (Q8) runs with room to spare (~81 tok/s est.). The cheapest card that also runs it comfortably is the RTX 3060 (12 GB, ~24 tok/s est.). Context ceiling on this card: 128k; on the RTX 3060: none.
Same engine as the verdict above: 90% of VRAM usable, KV-cache at fp16, bandwidth-derived tok/s. Verdicts of the other cards are computed for Qwen3.5 9B Instruct (Q8) Q8_0 exactly.
Quant explorer
Switch between the quality-gated builds tracked for this exact combo. Q4_K_M stays the default recommendation; heavier Q6/Q8 builds only appear when their exact Ollama tags are registry-verified.
Why no 2-bit builds: Q1/Q2-class quants can look attractive in a memory table, but their quality loss is large enough that ModelFit excludes them from rankings instead of inflating the catalog with junk options. Usable VRAM: 28.8 GB.
Workload verdicts
Qwen3.5 9B Instruct (Q8) on the RTX 5090, graded per use case from the model's registry-verified tuning and the engine's speed estimate for this exact combo.
Grades combine tag-verified tuning (a model not built for the workload caps at C) with the tok/s floor each workload needs to feel usable. A combo that partially offloads caps at C; one that does not fit is D everywhere.
Memory math: weights + context vs budget
Weights take 10.7 GB. Context costs extra KV-cache on top; this is where long-context sessions break on cards that technically fit the weights.
The model leaves about 18 GB of the usable VRAM budget free after weights and 16k context.
KV-cache figures assume an fp16 cache, the llama.cpp/Ollama default. Standard GQA models use a size-class estimate (8 KV heads x 128 head dim class); hybrid linear-attention models (Qwen3.5/3.6, Qwen3-Next) use the exact per-token cost from their published config, since only their sparse full-attention layers cache KV. A q8_0 KV cache roughly halves either figure. Estimates, not measurements.
See how fast it feels
A deterministic typing simulation for Qwen3.5 9B Instruct (Q8): first token ~0.4s (instant prefill), then ~81 tokens/sec.
ModelFit engine estimate, not a measured benchmark. Real speed varies with prompt length, thermals, runtime, and KV-cache settings.
Qwen3.5 9B Instruct (Q8) on RTX 5090: FAQ
Can the NVIDIA GeForce RTX 5090 run Qwen3.5 9B Instruct (Q8)?
Yes. Qwen3.5 9B Instruct (Q8) (Q8_0) loads in about 10.7 GB and the RTX 5090 offers 28.8 GB of usable VRAM, leaving headroom for context. Expect at roughly 81 tokens/sec (est.).
How much VRAM does Qwen3.5 9B Instruct (Q8) need?
About 10.7 GB for the weights at Q8_0, plus KV-cache for context: roughly 0.5 GB extra at 16k tokens. The RTX 5090 budget is 28.8 GB (32 GB x 90%).
What is the best quantization of Qwen3.5 9B Instruct (Q8) for the RTX 5090?
The Q4_K_M build is the highest quality that fits (7 GB vs 28.8 GB usable). The Q8_0 build at 10.7 GB leaves more room for long context.
What GPU do I need to run Qwen3.5 9B Instruct (Q8) comfortably?
The RTX 5090 already runs Qwen3.5 9B Instruct (Q8) comfortably. Larger cards only buy you longer context or a heavier quant.
How much context can Qwen3.5 9B Instruct (Q8) use on the RTX 5090?
Up to 128k tokens stay fully in VRAM (14.7 GB total with the KV cache). Quantizing the KV cache to q8_0 roughly halves the KV column and buys back about one context tier.
Is Qwen3.5 9B Instruct (Q8) on the RTX 5090 fast enough for coding agents?
Barely on paper: the engine estimates ~81 tok/s against the ~30 tok/s an agentic loop needs to feel responsive, but Qwen3.5 9B Instruct (Q8) is not tuned for tool-calling, which caps its agentic grade; a coding-tuned model is the better pick here.