Card × model benchmark

Coder V2 Lite 16B-A2.4B on RTX 4070 Ti Q4_K_M

Smooth

On an RTX 4070 Ti (12 GB), Coder V2 Lite 16B-A2.4B in Q4_K_M generates 189 tokens/s and reads the prompt at 7 068 tokens/s, using 10,6 Go of video memory.

Median of 1 run · latest on 25/09/26

189 tok/s

Generation

7 068 tok/s

Prompt processing

10,6 Go

VRAM used (peak)

96 W

Average power

1,97 tok/s/W

Efficiency

0,04 €/M tok

Electricity / M tokens · 0,25 €/kWh

Estimate your monthly cost from your electricity price and usage: sign in

RTX 4070 Ti page All cards on Coder V2 Lite 16B-A2.4B Feel the speed Rent in the cloud

Where to buyLDLC Materiel.net TopAchat Cybertek Grosbill Rue du Commerce Amazon

What about a longer context?

An estimate, not a measurement: the KV cache grows with context, takes VRAM and slows generation down.

Model architecture unknown: no estimate available yet.

Other models measured on the RTX 4070 Ti

Gemma 3 270M Q4_K_M 728 tok/s LFM2.5 1.2B Q4_K_M 473 tok/s Llama 3.2 1B Q4_K_M 442 tok/s R1 Distill 1.5B Qwen Q4_K_M 328 tok/s Gemma 2 2B Q4_K_M 202 tok/s Llama 3.2 3B Instruct Q4_K_M 189 tok/s Ministral 3 3B Q4_K_M 173 tok/s Granite 4.1 3B Q4_K_M 166 tok/s Phi-4 mini 3.8B Q4_K_M 159 tok/s Gemma 3n E2B Q4_K_M 163 tok/s Gemma 4 E2B Q4_K_M 198 tok/s Gemma 3n E4B Q4_K_M 107 tok/s Mistral 7B 7B Q4_K_M 98,1 tok/s Qwen2.5 Coder 7B Q4_K_M 96,7 tok/s Qwen2.5 7B Instruct Q4_K_M 98,8 tok/s R1 Distill 7B Qwen Q4_K_M 97,6 tok/s Meta Llama 3.1 8B Instruct Q4_K_M 91,9 tok/s R1 0528 (Qwen3) 8B Q4_K_M 90,1 tok/s Ministral 3 8B Q4_K_M 84,0 tok/s Gemma 4 E4B Q4_K_M 114 tok/s Gemma 2 9B Q4_K_M 72,1 tok/s Gemma 4 12B Q4_K_M 53,8 tok/s Ministral 3 14B Q4_K_M 53,8 tok/s Qwen2.5 14B Instruct Q4_K_M 50,5 tok/s DeepSeek R1 Distill Qwen 14B Q4_K_M 51,1 tok/s Qwen2.5 Coder 14B Q4_K_M 50,3 tok/s Phi-4 Q4_K_M 50,8 tok/s

All measurements

Date	Source	Type	Generation	Prompt	VRAM	Max temp.	Power	llama.cpp
25/09/26	SephiGame	context 8 192ASUS ROG Strix	0,7	77,3	11,6 Go	48 °C	66 W	llama.cpp 96278e39f
25/09/26	SephiGame	standardASUS ROG Strix	189	7 068	10,6 Go	58 °C	96 W	llama.cpp 96278e39f

Protocol: llama-bench (llama.cpp), prompt 512 + generation 128 tokens, model fully loaded in VRAM. VRAM used = system peak (includes the OS, ~0.5 to 1 GB more than the model alone). Measure my card