Can RTX 3070 8GB run Qwen3-Coder-Next?

NVIDIA · 8 GB GDDR6 · 448 GB/s bandwidth · Alibaba · 79.7B (active 3B) · MoE · model ctx up to 256K

No — Qwen3-Coder-Next at Q4 needs ≈ 45.14 GB even at 4K context, beyond the RTX 3070 8GB's 8 GB GDDR6. It would only run with heavy CPU/disk offload. FP16 needs ≈ 161.0 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	45.14 GB	No	—
Q4	8K default	45.44 GB	No	—
Q4	32K	47.24 GB	No	—
FP16	4K	160.7 GB	No	—
FP16	8K default	161.0 GB	No	—
FP16	32K	162.8 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 3070 8GB VRAM
Q4	43.84 GB	0.6 GB	1 GB	45.44 GB	8 GB
FP16	159.4 GB	0.6 GB	1 GB	161.0 GB	8 GB

Other GPUs that run Qwen3-Coder-Next

• B200 192GB SXM — ≈1697.0 tok/s Q4 est.
• MI300X 192GB — ≈1124.2 tok/s Q4 est.
• H200 141GB SXM5 — ≈1018.2 tok/s Q4 est.
• H100 80GB SXM5 — ≈710.6 tok/s Q4 est.
• Gaudi 3 128GB — ≈694.9 tok/s Q4 est.
• A100 80GB SXM4 — ≈432.5 tok/s Q4 est.
• RTX 6000 Ada 48GB — ≈203.6 tok/s Q4 est.
• L40S 48GB — ≈183.3 tok/s Q4 est.

All GPUs for Qwen3-Coder-Next →

Other models for the RTX 3070 8GB

• Qwen3-0.6B — ≈356.4 tok/s Q4 est.
• gemma-3-1b-it — ≈285.1 tok/s Q4 est.
• Llama-3.2-1B-Instruct — ≈237.6 tok/s Q4 est.
• Qwen3-1.7B — ≈142.5 tok/s Q4 est.
• Llama-3.2-3B-Instruct — ≈89.1 tok/s Q4 est.
• Phi-4-mini-instruct — ≈75.0 tok/s Q4 est.
• Phi-4-mini-reasoning — ≈75.0 tok/s Q4 est.
• gemma-3-4b-it — ≈73.1 tok/s Q4 est.

All models on the RTX 3070 8GB →