What LLMs can an RTX 5070 Ti 16GB run?

An RTX 5070 Ti 16GB gives a model 16 GB of memory and 896 GB/s of bandwidth. Of the 40 open models tracked here, 8 fit at Q4_K_M with an 8,192-token context, and 9 more at a lower precision.

Models that fit on one RTX 5070 Ti 16GB

The most precise weights that still fit with 8K tokens of context and one request, the memory that takes, the longest context at that precision and the writing speed. Bigger models come first.

Model	Best precision	Memory	Longest context	Tokens/s
Nemotron 3 Nano 30B-A3B 31.6B, 3.5B active	GGUF Q2_K	14.1 GB	256K (full)	140–257
Gemma 4 31B 31.3B	GGUF Q2_K	15.1 GB	28K	33–46
GLM-4.7 Flash 31.2B, 3B active	GGUF Q2_K	14.3 GB	36K	128–233
Xing 4.0 29B-A4B 31.2B, 4B active	GGUF Q2_K	14.3 GB	43K	109–197
Qwen3 30B-A3B 30.5B, 3.3B active	GGUF Q2_K	14.4 GB	23K	104–186
Muse Glimmer 30B 29.8B	GGUF Q3_K_M	15.6 GB	36K	32–45
Qwen3.6 27B 27.8B	GGUF Q3_K_M	15.0 GB	22K	33–47
Qwen3.8 27B 27.8B	GGUF Q3_K_M	15.0 GB	22K	33–47
Gemma 4 26B-A4B 25.8B, 4B active	GGUF Q3_K_M	13.7 GB	219K	101–181
gpt-oss-20b 20.9B, 3.6B active	GGUF Q5_K_M	15.9 GB	11K	85–150
Gemma 4 12B 12.0B	GGUF Q8_0	13.9 GB	248K	36–50
Qwen3.5 9B 9.7B	GGUF Q8_0	11.3 GB	145K	44–62
Qwen3 8B 8.2B	GGUF Q8_0	10.7 GB	40K (full)	46–66
Llama 3.1 8B 8.0B	GGUF Q8_0	10.3 GB	49K	48–68
Gemma 4 E4B 8.0B	GGUF Q8_0	9.36 GB	128K (full)	52–75
Nemotron 3 Nano 4B 4.0B	FP16 / BF16	8.78 GB	256K (full)	56–80
MiniCPM5 2B 2.5B	FP16 / BF16	6.02 GB	128K (full)	80–117

Too big for one card

How many RTX 5070 Ti 16GB cards these need at Q4_K_M in one tensor-parallel group; some need more than 8.

Qwen3.6 35B-A3B: 2 cards
Llama 3.1 70B: 4 cards
Qwen3-Coder-Next: 4 cards
gpt-oss-120b: 8 cards
Nemotron 3 Super 120B-A12B: 8 cards
Qwen3.5 122B-A10B: 8 cards
Mistral Medium 3.5 128B: 8 cards
Qwen3.8 Flash Next: 8 cards
MiniMax M2.7: more than 8 cards
DeepSeek V4 Flash: more than 8 cards
DeepSeek V4 Flash 0731: more than 8 cards
MiMo V2.6 Flash: more than 8 cards
GLM-5.3 Flash: more than 8 cards
MiniMax M3: more than 8 cards
DeepSeek V3 / R1: more than 8 cards
DeepSeek V3.2: more than 8 cards
GLM-5.3: more than 8 cards
GLM-5.2: more than 8 cards
DeepSeek V4.1 Flash: more than 8 cards
MiMo V2.6 Pro: more than 8 cards
DeepSeek V4 Pro: more than 8 cards
Qwen3.8 2.4T-A95B: more than 8 cards
Kimi K3: more than 8 cards

How these numbers are worked out

Memory is the weights at each precision plus the KV cache for 8,192 tokens in FP16 and runtime overhead (0.5 GB plus 10%), computed from each model's files on Hugging Face with the LLM VRAM Calculator. Speed is estimated from memory bandwidth and the parameters read per token with the LLM Speed Calculator. Both tools take any other model from Hugging Face.

Other GPUs and Macs

Model numbers read from Hugging Face on 2026-09-26.