AMD Radeon RX 7600 XT vs AMD Radeon RX 7800 XT

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)
AMD Radeon RX 7600 XT
  • 16GB
  • rdna3
  • amd
  • $329
  • 108 of 109
  • 0 models
AMD Radeon RX 7800 XT
  • 16GB
  • rdna3
  • amd
  • $499
  • 108 of 109
  • 0 models

Language Models47 models

F
—
Llama 3.1 70B Instruct
70B · Meta
—
F
C
—
Qwen 2.5 32B
32B · Alibaba
—
C
B
—
Gemma 3 27B
27B · Google
—
B
A
—
Mistral Small 22B
22B · Mistral AI
—
A
S
—
Phi-4
14B · Microsoft
26
S
S
—
Qwen 2.5 14B
14B · Alibaba
26
S
S
—
Gemma 3 12B
12B · Google
26
S
S
—
Mistral Nemo 12B
12B · Mistral AI
26
S
S
—
Solar 10.7B
10.7B · Upstage
26
S
S
22
Falcon 3 10B
10B · TII
42
S
S
22
Gemma 2 9B Instruct
9.2B · Google
42
S
S
22
Yi 1.5 9B Chat
9B · 01.AI
42
S
S
22
DeepSeek R1 Distill 8B
8B · DeepSeek
42
S
S
22
Llama 3.1 8B Instruct
8B · Meta
42
S
S
22
Granite 3.3 8B
8B · IBM
42
S
S
22
EXAONE 3.5 7.8B
7.8B · LG AI
42
S
S
22
InternLM 2.5 7B
7.7B · Shanghai AI Lab
42
S
S
22
Qwen 2.5 7B Instruct
7.6B · Alibaba
42
S
S
22
Mistral 7B Instruct v0.3
7.3B · Mistral AI
42
S
S
22
Falcon 3 7B
7B · TII
42
S
S
22
OLMo 2 7B
7B · Allen AI
42
S
S
22
OpenChat 3.5 7B
7B · OpenChat
42
S
S
22
Yi 1.5 6B Chat
6B · 01.AI
42
S
S
42
Gemma 3 4B
4B · Google
66
S
S
42
Nemotron Mini 4B
4B · NVIDIA
66
S
S
42
Danube 3 4B
4B · H2O.ai
66
S
S
42
Phi-3.5 Mini 3.8B
3.8B · Microsoft
66
S
S
42
Phi-4 Mini 3.8B
3.8B · Microsoft
66
S
S
42
Llama 3.2 3B Instruct
3.2B · Meta
66
S
S
42
Qwen 2.5 3B
3B · Alibaba
66
S
S
42
Falcon 3 3B
3B · TII
66
S
S
42
StableLM Zephyr 3B
3B · Stability AI
66
S
S
42
Rocket 3B
3B · Pansophic
66
S
S
42
Gemma 2 2B
2.6B · Google
66
S
S
42
EXAONE 3.5 2.4B
2.4B · LG AI
66
S
S
66
Granite 3.3 2B
2B · IBM
96
S
S
66
SmolLM2 1.7B
1.7B · HuggingFace
96
S
S
66
Qwen 2.5 1.5B
1.5B · Alibaba
96
S
S
66
DeepSeek R1 Distill 1.5B
1.5B · DeepSeek
96
S
S
66
Llama 3.2 1B Instruct
1.24B · Meta
96
S
S
66
TinyLlama 1.1B
1.1B · TinyLlama
96
S
S
66
Gemma 3 1B
1B · Google
96
S
S
66
Falcon 3 1B
1B · TII
96
S
S
66
Qwen 2.5 0.5B
0.5B · Alibaba
96
S
S
66
Danube 3 500M
0.5B · H2O.ai
96
S
S
66
SmolLM2 360M
0.36B · HuggingFace
96
S
S
66
SmolLM2 135M
0.135B · HuggingFace
96
S

Code Models16 models

S
—
Qwen 2.5 Coder 14B
14B · Alibaba
26
S
S
—
Code Llama 13B Instruct
13B · Meta
26
S
S
22
Yi Coder 9B
9B · 01.AI
42
S
S
22
CodeGemma 7B
8.5B · Google
42
S
S
22
Qwen 2.5 Coder 7B
7.6B · Alibaba
42
S
S
22
StarCoder2 7B
7B · BigCode
42
S
S
22
Code Llama 7B
7B · Meta
42
S
S
22
DeepSeek Coder 6.7B
6.7B · DeepSeek
42
S
S
42
Qwen 2.5 Coder 3B
3B · Alibaba
66
S
S
42
StarCoder2 3B
3B · BigCode
66
S
S
42
Stable Code 3B
3B · Stability AI
66
S
S
66
CodeGemma 2B
2B · Google
96
S
S
66
Qwen 2.5 Coder 1.5B
1.5B · Alibaba
96
S
S
66
Yi Coder 1.5B
1.5B · 01.AI
96
S
S
66
DeepSeek Coder 1.3B
1.3B · DeepSeek
96
S
S
66
Qwen 2.5 Coder 0.5B
0.5B · Alibaba
96
S

Multimodal & Vision6 models

S
22
LLaVA 1.6 7B
7B · LLaVA
42
S
S
42
Phi-3.5 Vision
4.2B · Microsoft
66
S
S
42
PaliGemma 3B
3B · Google
66
S
S
42
Qwen2-VL 2B
2.2B · Alibaba
66
S
S
66
MiniCPM-V 2.6
2B · OpenBMB
96
S
S
66
Moondream 2
1.8B · Moondream
96
S

Image Generation9 models

B
—
FLUX.1 Schnell (GGUF)
12B · Black Forest Labs
25
B
B
—
FLUX.1 Dev (GGUF)
12B · Black Forest Labs
25
B
S
42
Stable Diffusion XL (CoreML)
3.5B · Stability AI
66
S
S
42
SDXL Turbo (GGUF)
3.5B · Stability AI
66
S
S
42
Stable Diffusion 3 Medium (GGUF)
2.5B · Stability AI
66
S
S
66
Stable Diffusion 2.1 Base (CoreML)
0.86B · Stability AI / Apple
96
S
S
66
Stable Diffusion 1.5 (CoreML)
0.86B · Runway
96
S
S
66
Stable Diffusion 1.5 (GGUF)
0.86B · Runway / GPUStack
96
S
S
66
Stable Diffusion 2.1 (GGUF)
0.86B · Stability AI
96
S

Speech9 models

S
66
Whisper Large v3
1.55B · OpenAI
96
S
S
66
Whisper Large v3 Turbo
0.81B · OpenAI
96
S
S
66
Whisper Medium
0.77B · OpenAI
96
S
S
66
Distil-Whisper Large v3
0.76B · HuggingFace
96
S
S
66
Whisper Small
0.24B · OpenAI
96
S
S
66
Whisper Base
0.074B · OpenAI
96
S
S
66
Whisper Base English
0.074B · OpenAI
96
S
S
66
Whisper Tiny English (Quantized)
0.039B · OpenAI
96
S
S
66
Whisper Tiny
0.039B · OpenAI
96
S

Text-to-Speech14 models

S
66
Kokoro 82M TTS
0.082B · Kokoro
96
S
S
66
Piper TTS - Amy (English)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Lessac (English)
0.02B · Rhasspy
96
S
S
66
Piper TTS - LibriTTS-R (English)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Spanish (MLS)
0.02B · Rhasspy
96
S
S
66
Piper TTS - French (Siwis)
0.02B · Rhasspy
96
S
S
66
Piper TTS - German (Thorsten)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Chinese (Huayan)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Japanese (Kokoro)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Korean
0.02B · Rhasspy
96
S
S
66
Piper TTS - Russian (Irina)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Portuguese (Faber)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Italian (Riccardo)
0.02B · Rhasspy
96
S
S
66
Piper TTS - Arabic (Kareem)
0.02B · Rhasspy
96
S

Embeddings5 models

S
66
BGE Large EN v1.5
0.335B · BAAI
96
S
S
66
Nomic Embed Text v1.5
0.137B · Nomic AI
96
S
S
66
BGE Small EN v1.5
0.033B · BAAI
96
S
S
66
Snowflake Arctic Embed S
0.033B · Snowflake
96
S
S
66
all-MiniLM-L6-v2
0.023B · Sentence Transformers
96
S

Rerankers2 models

S
66
BGE Reranker v2 M3
0.568B · BAAI
96
S
S
66
Jina Reranker Tiny EN
0.033B · Jina AI
96
S