Hardware

GeForce RTX 5070

GeForce RTX 5070 has 12 GB GDDR7 and 672 GB/s. See which open models fit, at which quant, and roughly how fast. Figures below use the default memory (12 GB), 32 GB system RAM, and DDR5-5600 for anything that spills. Updated 2026-10-24.

Memory12 GB GDDR7

Bandwidth672 GB/s

Kindgpu

Year2025

NVIDIA official compare table: memory config and bandwidth. Source.

RSS for new models on this device

Models that fit

Model	Verdict	Quant	Context	tok/s
Qwen3 14B 14.8B	Tight fit	Q5_K_S	4K	34–46
NousResearch Hermes 4 14B 14.8B	Tight fit	Q5_K_S	4K	34–46
Ministral 3 14B Instruct 2512 13.5B	Tight fit	UD-Q5_K_XL	8K	34–46
Ministral 3 14B Reasoning 2512 13.5B	Tight fit	UD-Q5_K_XL	8K	34–46
Qwen2.5 14B Instruct 14.8B	Tight fit	Q4_K_M	8K	35–48
DeepSeek R1 Distill Qwen 14B 14.8B	Tight fit	Q4_K_M	8K	35–48
Qwen2.5 Coder 14B Instruct 128K 14.8B	Tight fit	Q4_K_M	8K	35–48
phi 4 14.7B	Tight fit	Q4_K_M	8K	35–48
gemma 4 12b it 11.9B	Tight fit	Q6_K	8K	28–39
Qwen3.6 14B A3B FableVibes 13.8B	Tight fit	Q5_K_M	8K	160–218
Devstral Small 2 24B Instruct 2512 23.6B	Tight fit	UD-IQ3_XXS	8K	34–47
Mistral Small 3.2 24B Instruct 2506 23.6B	Tight fit	UD-IQ3_XXS	8K	34–47
Magistral Small 2509 23.6B	Tight fit	UD-IQ3_XXS	8K	34–47
Qwen3.8 27B 27.3B	Tight fit	UD-Q2_K_XL	8K	36–49
Qwen2.5 32B Instruct 32.8B	Tight fit	IQ2_XXS	4K	37–50
Qwen2.5 Coder 32B Instruct 32.8B	Tight fit	IQ2_XXS	4K	37–50
Qwen3 32B 32.8B	Tight fit	UD-IQ2_XXS	4K	36–49
Qwen3 VL 32B Instruct 32.8B	Tight fit	UD-IQ2_XXS	4K	36–49
GLM 4 32B 0414 32.6B	Tight fit	UD-IQ2_XXS	8K	38–52
GLM 4.6V Flash 9.4B	Tight fit	Q8_0	8K	36–49
GLM Z1 9B 0414 9.4B	Tight fit	Q8_0	8K	36–49
GLM 4 9B 0414 9.4B	Tight fit	Q8_0	8K	36–49
gemma 4 31B it 30.7B	Tight fit	UD-IQ2_XXS	8K	22–30
Qwen3.8 9B Distill 9.2B	Tight fit	Q8_0	8K	37–50
Qwen3.5 9B 9.0B	Tight fit	Q8_0	8K	38–51
MiMo V2.6 Distill Qwen 9B 9.0B	Tight fit	Q8_0	8K	38–51
Qwythos 9B Claude Mythos 5 1M 9.0B	Tight fit	Q8_0	8K	38–51
Qwythos 9B v2 9.0B	Tight fit	Q8_0	8K	38–51
granite 4.2 8b 8.8B	Tight fit	Q8_0	8K	35–47
Ministral 3 8B Instruct 2512 8.5B	Tight fit	Q8_0	8K	36–50
Ministral 3 8B Reasoning 2512 8.5B	Tight fit	Q8_0	8K	36–50
Swift Qwen3.8 27B 27.3B	Tight fit	IQ2_S	8K	35–48
Qwen3.5 27B 26.9B	Runs	UD-IQ2_XXS	8K	41–55
Qwen3.6 27B 26.9B	Tight fit	UD-IQ2_XXS	8K	37–51
Qwen3 VL 8B Instruct 8.2B	Tight fit	Q8_0	8K	37–51
DeepSeek R1 0528 Qwen3 8B 8.2B	Tight fit	Q8_0	8K	37–51
Qwen3 8B 8.2B	Tight fit	Q8_0	8K	37–51
Qwen3 VL 8B Thinking 8.2B	Tight fit	Q8_0	8K	37–51
gemma 3 27b it 27.0B	Runs	UD-IQ2_XXS	8K	31–42
medgemma 27b text it 27.0B	Runs	UD-IQ2_XXS	8K	31–42

105 more models in the catalog do not fit this default profile, or ranked below the first 40 that do.

Example: Tight fit fully on your GeForce RTX 5070 at Q5_K_S, 4K context, about 34–46 tok/s. Weights 9.56 GB plus 0.63 GB of KV cache — 10.6 GB of 10.9 GB usable. Weights shown in the verdict use GiB (the same 1024³ counting as GPU memory).