readygpu

NVIDIA GeForce RTX 4060 Ti 8GB vs NVIDIA GeForce RTX 5090

For running models locally, priced in Canada. At 8k context the first runs 26 models fully in memory, the second 62.

In short

The NVIDIA GeForce RTX 5090 has more memory for models: 32 GB against 8 GB. At 8k context the NVIDIA GeForce RTX 5090 runs 62 of the models we track fully in memory and the NVIDIA GeForce RTX 4060 Ti 8GB 26; only the NVIDIA GeForce RTX 5090 holds Qwen 3.8 27B. On the models both run fully in memory, the NVIDIA GeForce RTX 5090 is about 6.2× as fast as the NVIDIA GeForce RTX 4060 Ti 8GB (estimated, median across models). In Canada the NVIDIA GeForce RTX 4060 Ti 8GB costs $864 at Amazon and the NVIDIA GeForce RTX 5090 $10,101 at Amazon: the NVIDIA GeForce RTX 4060 Ti 8GB is $9,237 cheaper.

Updated . Speeds are estimates; how they are worked out.

NVIDIA GeForce RTX 4060 Ti 8GB

$864

Amazon

Add to build
NVIDIA GeForce RTX 5090

$10,101

Amazon

Add to build

Specifications

SpecNVIDIA GeForce RTX 4060 Ti 8GBNVIDIA GeForce RTX 5090
GPUGeForce RTX 4060 Ti 8GBGeForce RTX 5090
Memory8 GB GDDR632 GB GDDR7
Memory bus128-bit512-bit
Memory bandwidth288 GB/s1,792 GB/s
FP16 compute22.065 TFLOPS209 TFLOPS
SoftwareCUDACUDA
Board power160 W575 W
PCIe generationGen 4Gen 5
PCIe lanesx8x16

What each runs

Estimated decode speed with the most precise quant that fits each one est.

ModelNVIDIA GeForce RTX 4060 Ti 8GBNVIDIA GeForce RTX 5090
Qwen 3.8 27Boffload37tok/sQ8_0
Qwen 3.6 27Boffload38tok/sQ8_0
Gemma 4 31Boffload42tok/sQ6_K
Qwen 3.6 35B-A3Boffload237tok/sQ6_K
Qwen 3.5 27Boffload38tok/sQ8_0
Gemma 4 26B-A4Boffload147tok/sQ8_0
Qwen 3.5 35B-A3Boffload253tok/sQ6_K
GLM 4.7 Flashoffload117tok/sQ8_0
Devstral Small 2 24Boffload42tok/sQ8_0
Nemotron 3.5 Lightning 30B-A3Boffload225tok/sUD-Q5_K_M
gpt-oss 20Boffload253tok/sMXFP4
Kimi Linear 48B-A3Boffload343tok/sQ4_K_M
Qwen3 Coder 30B-A3Boffload204tok/sQ6_K
Qwen3 32Boffload38tok/sQ6_K
Gemma 4 12Boffload85tok/sQ8_0
Qwen 3.5 9B26tok/sQ5_K_M113tok/sQ8_0
Mistral Small 3.2 24Boffload42tok/sQ8_0
Gemma 3 27Boffload37tok/sQ8_0
DeepSeek R1 Distill Qwen 32Boffload38tok/sQ6_K
Ministral 3 14Boffload71tok/sQ8_0
Qwen3 14Boffload65tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B25tok/sQ5_K_M112tok/sQ8_0
Qwen3 8B25tok/sQ5_K_M112tok/sQ8_0
Phi-4 14Boffload64tok/sQ8_0
Qwen 3.5 4B37tok/sQ8_0231tok/sQ8_0
Ministral 3 8B28tok/sQ4_K_M109tok/sQ8_0
Granite 4.2 8B26tok/sQ4_K_M104tok/sQ8_0
Gemma 4 E4B31tok/sQ5_K_M131tok/sQ8_0
LFM2.5 8B-A1B93tok/sUD-Q5_K_M455tok/sQ8_0
Llama 3.1 8B26tok/sQ5_K_M145tok/sQ6_K
Qwen2.5 0.5B283tok/sQ8_01759tok/sQ8_0
Qwen3.5 0.8B192tok/sQ8_01193tok/sQ8_0
Qwen3.5 2B115tok/sQ5_K_M715tok/sQ5_K_M
LFM3 1.7B91tok/sQ8_0568tok/sQ8_0
DeepSeek R1 Distill Qwen 1.5B84tok/sQ8_0522tok/sQ8_0
MiniCPM5 2B59tok/sQ8_0366tok/sQ8_0
LFM2.5 2.6B53tok/sQ8_0329tok/sQ8_0
LFM2.5 VL 3B53tok/sQ8_0329tok/sQ8_0
Spark X2.5 4B53tok/smlx-4bit327tok/smlx-4bit
granite vision 4.1 4b42tok/sQ8_0259tok/sQ8_0
Ministral 3 3B Instruct 251240tok/sQ8_0246tok/sQ8_0
granite 4.2 3b39tok/sQ8_0243tok/sQ8_0
FrogNano 4B 260936tok/sQ8_0225tok/sQ8_0
Ministral 3 8B Reasoning 251228tok/sQ4_K_M109tok/sQ8_0
GLM 4.6V Flash28tok/sQ4_K_M108tok/sQ8_0
DeepSeek R1 Distill Qwen 7B27tok/sQ6_K130tok/sQ8_0
DeepSeek R1 Distill Llama 8B26tok/sQ5_K_M116tok/sQ8_0
Mellum2.1 12B A2.5B Thinkingoffload236tok/sQ8_0
Kimi VL A3B Thinking 2506offload199tok/sQ8_0
LFM2.5 24B A2Boffload238tok/sQ8_0
Kimi VL A3Boffload190tok/smlx-8bit
GLM 4 9B 0414offload129tok/smlx-5bit
Ministral 3 14B Reasoning 2512offload71tok/sQ8_0
DeepSeek R1 Distill Qwen 14Boffload64tok/sQ8_0
GLM 4.1V 9B Thinkingoffload92tok/smlx-8bit
Devstral Small 2507offload42tok/sQ8_0
Magistral Small 2506offload42tok/sQ8_0
Magistral Small 2507offload42tok/sQ8_0
Magistral Small 2509offload42tok/sQ8_0
Muse Glimmer 30Boffload34tok/sQ8_0
granite 4.2 30boffload42tok/sQ6_K
Qwen3.8 27B OBLITERATEDoffload52tok/smlx-4bit
Qwen 3.5 122B-A10Bwon't runoffload
Nemotron 3 Super 120B-A12Bwon't runoffload
Qwen3 Coder Nextwon't runoffload
Mistral Small 4 119Bwon't runoffload
GLM 4.5 Airwon't runoffload
Qwen3 Next 80B-A3Bwon't runoffload
Llama 3.3 70Bwon't runoffload
DeepSeek R1 Distill Llama 70Bwon't runoffload
DeepSeek V4 Flashwon't runwon't run
GLM 5.3 Flashwon't runwon't run
Qwen 3.8 Flash Nextwon't runwon't run
MiniMax M2.7won't runwon't run
gpt-oss 120Bwon't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek R1 Zerowon't runwon't run
DeepSeek V3won't runwon't run
DeepSeek V3 0324won't runwon't run
DeepSeek V3.1won't runwon't run
DeepSeek V3.1 Terminuswon't runwon't run
DeepSeek V3.2won't runwon't run
DeepSeek V4 Flash 0731won't runwon't run
DeepSeek V4 Flash Vision Expwon't runwon't run
Devstral 2 123B Instruct 2512won't runwon't run
GLM 4.5won't runwon't run
GLM 4.5Vwon't runwon't run
GLM 4.6won't runwon't run
GLM 4.6Vwon't runwon't run
GLM 4.7won't runwon't run
GLM 5won't runwon't run
GLM 5.1won't runwon't run
GLM 5.2won't runwon't run
GLM 5.3won't runwon't run
Kimi Dev 72Bwon't runwon't run
Kimi K2won't runwon't run
Kimi K2 Instruct 0905won't runwon't run
Kimi K2 Thinkingwon't runwon't run
Kimi K2.5won't runwon't run
MiniMax M2won't runwon't run
MiniMax M2.1won't runwon't run
MiniMax M2.5won't runwon't run
MiniMax M3won't runwon't run
Mistral Medium 3.5 128Bwon't runwon't run

Questions

Which is better for local AI, the NVIDIA GeForce RTX 4060 Ti 8GB or the NVIDIA GeForce RTX 5090?
Choose the NVIDIA GeForce RTX 5090 to run bigger models. Choose the NVIDIA GeForce RTX 5090 for speed. Choose the NVIDIA GeForce RTX 4060 Ti 8GB to spend less.
Which can run bigger models, the NVIDIA GeForce RTX 4060 Ti 8GB or the NVIDIA GeForce RTX 5090?
At 8k context the NVIDIA GeForce RTX 5090 runs 62 of the models we track fully in memory and the NVIDIA GeForce RTX 4060 Ti 8GB 26; only the NVIDIA GeForce RTX 5090 holds Qwen 3.8 27B.
Which is faster for local AI, the NVIDIA GeForce RTX 4060 Ti 8GB or the NVIDIA GeForce RTX 5090?
On the models both run fully in memory, the NVIDIA GeForce RTX 5090 is about 6.2× as fast as the NVIDIA GeForce RTX 4060 Ti 8GB (estimated, median across models).
What is the price difference between the NVIDIA GeForce RTX 4060 Ti 8GB and the NVIDIA GeForce RTX 5090?
In Canada the NVIDIA GeForce RTX 4060 Ti 8GB costs $864 at Amazon and the NVIDIA GeForce RTX 5090 $10,101 at Amazon: the NVIDIA GeForce RTX 4060 Ti 8GB is $9,237 cheaper.

More comparisons

More hardware for local AI