Can RTX 3080 10GB run Kimi-K2-Thinking?

NVIDIA · 10 GB GDDR6X · 760 GB/s bandwidth · Moonshot · 1058.1B (active 32B) · MoE · model ctx up to 256K

No — Kimi-K2-Thinking at Q4 needs ≈ 586.16 GB even at 4K context, beyond the RTX 3080 10GB's 10 GB GDDR6X. It would only run with heavy CPU/disk offload. FP16 needs ≈ 2123.6 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	586.16 GB	No	—
Q4	8K default	589.36 GB	No	—
Q4	32K	608.56 GB	No	—
FP16	4K	2120.4 GB	No	—
FP16	8K default	2123.6 GB	No	—
FP16	32K	2142.8 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 3080 10GB VRAM
Q4	581.96 GB	6.4 GB	1 GB	589.36 GB	10 GB
FP16	2116.2 GB	6.4 GB	1 GB	2123.6 GB	10 GB

Other GPUs that run Kimi-K2-Thinking

No catalog GPU fully fits Kimi-K2-Thinking at Q4/8K — see the model page for offload estimates.

All GPUs for Kimi-K2-Thinking →

Other models for the RTX 3080 10GB

• Qwen3-0.6B — ≈604.5 tok/s Q4 est.
• gemma-3-1b-it — ≈483.6 tok/s Q4 est.
• Llama-3.2-1B-Instruct — ≈403.0 tok/s Q4 est.
• Qwen3-1.7B — ≈241.8 tok/s Q4 est.
• Llama-3.2-3B-Instruct — ≈151.1 tok/s Q4 est.
• Phi-4-mini-instruct — ≈127.3 tok/s Q4 est.
• Phi-4-mini-reasoning — ≈127.3 tok/s Q4 est.
• gemma-3-4b-it — ≈124.0 tok/s Q4 est.

All models on the RTX 3080 10GB →