NVIDIA GeForce RTX 4070 Ti vs NVIDIA GeForce RTX 4070 Ti SUPER
Head-to-head AI inference comparison across 109 popular models. Each model is graded against both cards using its highest-quality quantization that still fits in VRAM. Bigger grade and faster tokens-per-second wins.
Spec
- VRAM
- Architecture
- Vendor
- MSRP
- Models running
- Wins (grade)
NVIDIA GeForce RTX 4070 Ti
- 12GB
- ada lovelace
- nvidia
- $799
- 108 of 109
- 0 models
NVIDIA GeForce RTX 4070 Ti SUPER
- 16GB
- ada lovelace
- nvidia
- $799
- 108 of 109
- 9 models
Where NVIDIA GeForce RTX 4070 Ti pulls ahead
- No standout wins.
Where NVIDIA GeForce RTX 4070 Ti SUPER pulls ahead
- Phi-4AvsS
- Qwen 2.5 14BAvsS
- Qwen 2.5 32BDvsC
- Qwen 2.5 Coder 14BAvsS
- Stable Diffusion 3 Medium (GGUF)AvsS
Language Models47 models
Ti
tok/s
Model
tok/s
SUPER
D
—
Gemma 3 27B
27B · Google
—
B
C
—
Mistral Small 22B
22B · Mistral AI
—
A
D
—
Qwen 2.5 32B
32B · Alibaba
—
C
A
36
Phi-4
14B · Microsoft
42
S
A
36
Qwen 2.5 14B
14B · Alibaba
42
S
F
—
Llama 3.1 70B Instruct
70B · Meta
—
F
S
36
Gemma 3 12B
12B · Google
42
S
S
36
Mistral Nemo 12B
12B · Mistral AI
42
S
S
36
Solar 10.7B
10.7B · Upstage
42
S
S
62
Falcon 3 10B
10B · TII
70
S
S
62
Gemma 2 9B Instruct
9.2B · Google
70
S
S
62
Yi 1.5 9B Chat
9B · 01.AI
70
S
S
62
DeepSeek R1 Distill 8B
8B · DeepSeek
70
S
S
62
Llama 3.1 8B Instruct
8B · Meta
70
S
S
62
Granite 3.3 8B
8B · IBM
70
S
S
62
EXAONE 3.5 7.8B
7.8B · LG AI
70
S
S
62
InternLM 2.5 7B
7.7B · Shanghai AI Lab
70
S
S
62
Qwen 2.5 7B Instruct
7.6B · Alibaba
70
S
S
62
Mistral 7B Instruct v0.3
7.3B · Mistral AI
70
S
S
62
Falcon 3 7B
7B · TII
70
S
S
62
OLMo 2 7B
7B · Allen AI
70
S
S
62
OpenChat 3.5 7B
7B · OpenChat
70
S
S
62
Yi 1.5 6B Chat
6B · 01.AI
70
S
S
94
Gemma 3 4B
4B · Google
102
S
S
94
Nemotron Mini 4B
4B · NVIDIA
102
S
S
94
Danube 3 4B
4B · H2O.ai
102
S
S
94
Phi-3.5 Mini 3.8B
3.8B · Microsoft
102
S
S
94
Phi-4 Mini 3.8B
3.8B · Microsoft
102
S
S
94
Llama 3.2 3B Instruct
3.2B · Meta
102
S
S
94
Qwen 2.5 3B
3B · Alibaba
102
S
S
94
Falcon 3 3B
3B · TII
102
S
S
94
StableLM Zephyr 3B
3B · Stability AI
102
S
S
94
Rocket 3B
3B · Pansophic
102
S
S
94
Gemma 2 2B
2.6B · Google
102
S
S
94
EXAONE 3.5 2.4B
2.4B · LG AI
102
S
S
132
Granite 3.3 2B
2B · IBM
144
S
S
132
SmolLM2 1.7B
1.7B · HuggingFace
144
S
S
132
Qwen 2.5 1.5B
1.5B · Alibaba
144
S
S
132
DeepSeek R1 Distill 1.5B
1.5B · DeepSeek
144
S
S
132
Llama 3.2 1B Instruct
1.24B · Meta
144
S
S
132
TinyLlama 1.1B
1.1B · TinyLlama
144
S
S
132
Gemma 3 1B
1B · Google
144
S
S
132
Falcon 3 1B
1B · TII
144
S
S
132
Qwen 2.5 0.5B
0.5B · Alibaba
144
S
S
132
Danube 3 500M
0.5B · H2O.ai
144
S
S
132
SmolLM2 360M
0.36B · HuggingFace
144
S
S
132
SmolLM2 135M
0.135B · HuggingFace
144
S
Code Models16 models
Ti
tok/s
Model
tok/s
SUPER
A
36
Qwen 2.5 Coder 14B
14B · Alibaba
42
S
S
36
Code Llama 13B Instruct
13B · Meta
42
S
S
62
Yi Coder 9B
9B · 01.AI
70
S
S
62
CodeGemma 7B
8.5B · Google
70
S
S
62
Qwen 2.5 Coder 7B
7.6B · Alibaba
70
S
S
62
StarCoder2 7B
7B · BigCode
70
S
S
62
Code Llama 7B
7B · Meta
70
S
S
62
DeepSeek Coder 6.7B
6.7B · DeepSeek
70
S
S
94
Qwen 2.5 Coder 3B
3B · Alibaba
102
S
S
94
StarCoder2 3B
3B · BigCode
102
S
S
94
Stable Code 3B
3B · Stability AI
102
S
S
132
CodeGemma 2B
2B · Google
144
S
S
132
Qwen 2.5 Coder 1.5B
1.5B · Alibaba
144
S
S
132
Yi Coder 1.5B
1.5B · 01.AI
144
S
S
132
DeepSeek Coder 1.3B
1.3B · DeepSeek
144
S
S
132
Qwen 2.5 Coder 0.5B
0.5B · Alibaba
144
S
Multimodal & Vision6 models
Ti
tok/s
Model
tok/s
SUPER
Image Generation9 models
Ti
tok/s
Model
tok/s
SUPER
C
26
FLUX.1 Schnell (GGUF)
12B · Black Forest Labs
40
B
C
26
FLUX.1 Dev (GGUF)
12B · Black Forest Labs
40
B
A
94
Stable Diffusion 3 Medium (GGUF)
2.5B · Stability AI
102
S
S
94
Stable Diffusion XL (CoreML)
3.5B · Stability AI
102
S
S
94
SDXL Turbo (GGUF)
3.5B · Stability AI
102
S
S
132
Stable Diffusion 2.1 Base (CoreML)
0.86B · Stability AI / Apple
144
S
S
132
Stable Diffusion 1.5 (CoreML)
0.86B · Runway
144
S
S
132
Stable Diffusion 1.5 (GGUF)
0.86B · Runway / GPUStack
144
S
S
132
Stable Diffusion 2.1 (GGUF)
0.86B · Stability AI
144
S
Speech9 models
Ti
tok/s
Model
tok/s
SUPER
S
132
Whisper Large v3
1.55B · OpenAI
144
S
S
132
Whisper Large v3 Turbo
0.81B · OpenAI
144
S
S
132
Whisper Medium
0.77B · OpenAI
144
S
S
132
Distil-Whisper Large v3
0.76B · HuggingFace
144
S
S
132
Whisper Small
0.24B · OpenAI
144
S
S
132
Whisper Base
0.074B · OpenAI
144
S
S
132
Whisper Base English
0.074B · OpenAI
144
S
S
132
Whisper Tiny English (Quantized)
0.039B · OpenAI
144
S
S
132
Whisper Tiny
0.039B · OpenAI
144
S
Text-to-Speech14 models
Ti
tok/s
Model
tok/s
SUPER
S
132
Kokoro 82M TTS
0.082B · Kokoro
144
S
S
132
Piper TTS - Amy (English)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Lessac (English)
0.02B · Rhasspy
144
S
S
132
Piper TTS - LibriTTS-R (English)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Spanish (MLS)
0.02B · Rhasspy
144
S
S
132
Piper TTS - French (Siwis)
0.02B · Rhasspy
144
S
S
132
Piper TTS - German (Thorsten)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Chinese (Huayan)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Japanese (Kokoro)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Korean
0.02B · Rhasspy
144
S
S
132
Piper TTS - Russian (Irina)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Portuguese (Faber)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Italian (Riccardo)
0.02B · Rhasspy
144
S
S
132
Piper TTS - Arabic (Kareem)
0.02B · Rhasspy
144
S
Embeddings5 models
Ti
tok/s
Model
tok/s
SUPER
Rerankers2 models
Ti
tok/s
Model
tok/s
SUPER