Can RTX 4090D 24GB (China) run Mistral-Small-4-119B-2603?

NVIDIA · 24 GB GDDR6X · 1008 GB/s bandwidth · Mistral · 119.0B (active 6.5B) · MoE · model ctx up to 256K

No — Mistral-Small-4-119B-2603 at Q4 needs ≈ 67.1 GB even at 4K context, beyond the RTX 4090D 24GB (China)'s 24 GB GDDR6X. It would only run with heavy CPU/disk offload. Decode at Q4/8K: ≈ 35.4 tok/s (estimated, batch 1). FP16 needs ≈ 240.3 GB — does not fit on this card. Estimates come from memory-bandwidth math; rows tagged measured override estimates. Relative ranking is more reliable than absolute tok/s. Methodology.

Fit & speed by quant and context

Quant	Context	VRAM needed	Fits?	tok/s (decode)
Q4	4K	67.1 GB	No	36.1 est.
Q4	8K default	67.75 GB	No	35.4 est.
Q4	32K	71.65 GB	No	31.6 est.
FP16	4K	239.65 GB	No	—
FP16	8K default	240.3 GB	No	—
FP16	32K	244.2 GB	No	—

Fits = weights + KV(ctx) + 1 GB overhead ≤ 95% of VRAM. Measured anchors are context-agnostic; the fit verdict is recomputed per context. A missing tok/s means the model is far beyond this card (offload-only territory).

VRAM breakdown at 8K context

Quant	Weights	KV cache	Overhead	Total needed	RTX 4090D 24GB (China) VRAM
Q4	65.45 GB	1.3 GB	1 GB	67.75 GB	24 GB
FP16	238.0 GB	1.3 GB	1 GB	240.3 GB	24 GB

Other GPUs that run Mistral-Small-4-119B-2603

• B200 192GB SXM — ≈783.2 tok/s Q4 est.
• MI300X 192GB — ≈518.9 tok/s Q4 est.
• H200 141GB SXM5 — ≈469.9 tok/s Q4 est.
• H100 80GB SXM5 — ≈328.0 tok/s Q4 est.
• Gaudi 3 128GB — ≈320.7 tok/s Q4 est.
• A100 80GB SXM4 — ≈199.6 tok/s Q4 est.
• Mac Studio M3 Ultra 96GB — ≈80.2 tok/s Q4 est.
• Mac Studio M2 Ultra 192GB — ≈78.3 tok/s Q4 est.

All GPUs for Mistral-Small-4-119B-2603 →

Other models for the RTX 4090D 24GB (China)

• MiniCPM5-1B — ≈593.9 tok/s Q4 est.
• Qwen3.5-2B — ≈320.7 tok/s Q4 est.
• GLM-4.7-Flash — ≈213.8 tok/s Q4 est.
• Qwen3-30B-A3B-Instruct-2507 — ≈213.8 tok/s Q4 est.
• Qwen3.5-35B-A3B — ≈213.8 tok/s Q4 est.
• Qwen3.6-35B-A3B — ≈213.8 tok/s Q4 est.
• Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 — ≈213.8 tok/s Q4 est.
• Ornith-1.0-35B — ≈213.8 tok/s Q4 est.

All models on the RTX 4090D 24GB (China) →