NVIDIA · consumer

GeForce RTX 5060 Ti

GeForce RTX 5060 Ti has 8 GB of VRAM at 448 GB/s — about 7.44 GiB usable after driver and compositor overhead. 1400 of 2118 indexed models fit at 4K context with f16 KV.

Spec sheet· bandwidth, theoreticalFrom the file· fit from summed bytesPredicted· speed
Memory
8 GB
GDDR7
Bandwidth
448 GB/s
128-bit bus
Tensor FP16
95 TF
dense
TDP
180 W
$379 MSRP
Context4K8K16K32K64K128K
KV cachef16q8_0q4_0quantizing the KV cache is a ~2× lever on the dominant term at long context
text 1204vision language 101audio tts 21video 8image 2embedding 26audio asr 38

What fits at 4K context

largest quantization that fits, per model · 1400 of 2118 indexed
ModelBest quantParamsWeights●KV●Total in memory◐Headroom◐tok/s◐
DeepSeek-R1-Distill-Llama-8B-AbliteratedI1-IQ3_XXS8.0B6.10 GiB0.50 GiB7.44 GiB0.00 GiB48±12.9%
OmniAtlas-Qwen3-30B-A3BI1-IQ1_M31.7B6.59 GiB0.00 GiB7.44 GiB0.00 GiB48±12.9%
Qwen3-Omni-30B-A3B-CaptionerI1-IQ1_M31.7B6.59 GiB0.00 GiB7.44 GiB0.00 GiB48±12.9%
gemma-4-12BQ3_K_M12.0B5.87 GiB0.72 GiB7.43 GiB0.01 GiB48±12.9%
Marco-Mini-InstructMoEI1-IQ3_XXS17.3B6.22 GiB0.44 GiB7.43 GiB0.01 GiB166±37%
Grug-12BQ3_K_M12.0B5.87 GiB0.72 GiB7.43 GiB0.01 GiB48±12.9%
gemma-4-12B-it-Esper4Q3_K_M12.0B5.87 GiB0.72 GiB7.43 GiB0.01 GiB48±12.9%
gemma-4-12B-itQ3_K_M12.0B5.87 GiB0.72 GiB7.43 GiB0.01 GiB48±12.9%
Fimbulvetr-11B-v2I1-Q4_K_S10.7B5.84 GiB0.75 GiB7.43 GiB0.01 GiB48±12.9%
SambaLingo-Japanese-ChatI1-Q5_K_M6.9B4.60 GiB2.00 GiB7.43 GiB0.01 GiB48±12.9%
NousCoder-14BIQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
spoomplesmaxx-mini-14BI1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
vanilla-cn-roleplay-0.2I1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Claria-14bI1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
NTX-2.1-ProI1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Qwen3-14B-UncensoredI1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
FrogMini-14B-2510I1-IQ3_XS—5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Qwen3-14B-abliteratedIQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Josiefied-Qwen3-14B-abliterated-v3IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Hermes-4-14BIQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Slava-Qwen3-14B-SerbianI1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Huihui-Qwen3-14B-abliterated-v2I1-IQ3_XS14.8B5.94 GiB0.63 GiB7.42 GiB0.02 GiB48±12.9%
Apriel-1.6-15b-ThinkerQ2_K_L14.9B5.82 GiB0.75 GiB7.42 GiB0.02 GiB48±12.9%
AMALIA-9B-0626-DPOQ5_K_S9.2B5.93 GiB0.66 GiB7.42 GiB0.02 GiB48±12.9%
Ministral-3-8B-Instruct-2512-BF16Q5_K_M8.9B6.04 GiB0.53 GiB7.42 GiB0.02 GiB48±12.9%
DeepSeek-Coder-V2-Lite-BaseMoEI1-IQ3_XXS15.7B6.49 GiB0.12 GiB7.41 GiB0.03 GiB154±37%
DeepSeek-Coder-V2-Lite-InstructMoEIQ3_XXS15.7B6.49 GiB0.12 GiB7.41 GiB0.03 GiB154±37%
DeepSeek-V2-Lite-ChatMoEIQ3_XXS15.7B6.49 GiB0.12 GiB7.41 GiB0.03 GiB154±37%
Qwen3.5-27B-Engineer-Deckard-GeminiI1-IQ1_M27.7B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Qwen3.5-27B-HERETIC-Polaris-Advanced-Thinking-Alpha-uncensoredI1-IQ1_M27.4B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Qwen3.5-27B-Deckard-PKD-Heretic-Uncensored-ThinkingI1-IQ1_M27.4B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Huihui-Qwen3.5-27B-abliteratedI1-IQ1_M27.8B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Qwen3.5-27B-Unredacted-MAXI1-IQ1_M27.4B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Qwen3.5-27B-hereticI1-IQ1_M27.4B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Qwen3.5-27B-DerestrictedI1-IQ1_M27.8B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-DistilledI1-IQ1_M27.8B6.30 GiB0.25 GiB7.41 GiB0.03 GiB48±12.9%
granite-4.0-7B-A1B-Creative-v0.1MoEQ8_06.7B6.62 GiB0.03 GiB7.41 GiB0.03 GiB161±37%
glm-4v-9bQ5_K_M13.9B6.57 GiB0.00 GiB7.41 GiB0.03 GiB48±12.9%
GLM-4.7-Flash-DerestrictedMoEI1-IQ1_M31.2B6.39 GiB0.21 GiB7.41 GiB0.03 GiB170±37%
Huihui-GLM-4.7-Flash-abliteratedMoEI1-IQ1_M31.2B6.39 GiB0.21 GiB7.41 GiB0.03 GiB170±37%
orpheus-3b-0.1-ftF163.8B6.16 GiB0.44 GiB7.40 GiB0.04 GiB48±12.9%
Qwen3-Coder-REAP-25B-A3BMoEIQ2_XXS24.9B6.24 GiB0.38 GiB7.40 GiB0.04 GiB140±37%
GLM-4.7-Flash-REAP-23B-A3B-absolute-heresyMoEI1-IQ2_XS23.0B6.38 GiB0.21 GiB7.40 GiB0.04 GiB153±37%
Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-HightopQ4_012.1B6.43 GiB0.09 GiB7.39 GiB0.05 GiB49±12.9%
InternVL3_5-8BQ6_K_L8.5B6.54 GiB0.00 GiB7.39 GiB0.05 GiB48±12.9%
HunyuanVideo-1.5Q6_K8.3B6.54 GiB0.00 GiB7.39 GiB0.05 GiB49±12.9%
Qwen2.5-3B-Instruct-abliteratedQ8_03.1B6.43 GiB0.14 GiB7.38 GiB0.06 GiB48±12.9%
Qwen3-15B-A2B-BaseMoEQ3_K_S15.6B6.37 GiB0.19 GiB7.36 GiB0.08 GiB173±37%
Mistral-7B-v0.1KV unresolvedQ4_K_M7.2B6.02 GiB0.50 GiB7.35 GiB0.09 GiB49±12.9%
MathCoder2-CodeLlama-7BQ5_K_L6.7B4.53 GiB2.00 GiB7.35 GiB0.09 GiB49±12.9%
Mistral-NeMo-Minitron-8B-InstructQ5_K_L8.4B5.90 GiB0.63 GiB7.35 GiB0.09 GiB49±12.9%
Phi-3-mini-4k-instructKV unresolvedIQ3_XXS3.8B5.05 GiB1.50 GiB7.35 GiB0.09 GiB48±12.9%
Falcon3-10B-InstructQ4_K_M10.3B5.86 GiB0.63 GiB7.35 GiB0.09 GiB49±12.9%
LFM2-24B-A2BMoEIQ2_S23.8B6.45 GiB0.08 GiB7.34 GiB0.10 GiB191±37%
Fara1.5-9BQ5_K_M9.4B6.38 GiB0.13 GiB7.34 GiB0.10 GiB49±12.9%
QwenPaw-Flash-9BQ5_K_M9.4B6.38 GiB0.13 GiB7.34 GiB0.10 GiB49±12.9%
grug-9bQ5_K_M9.4B6.38 GiB0.13 GiB7.34 GiB0.10 GiB49±12.9%
OmniCoder-9BQ5_K_M9.4B6.38 GiB0.13 GiB7.34 GiB0.10 GiB49±12.9%
Ornith-1.0-9BQ5_K_M9.2B6.38 GiB0.13 GiB7.34 GiB0.10 GiB49±12.9%
Qwen3.5-9B-NeoQ5_K_M9.7B6.38 GiB0.13 GiB7.34 GiB0.10 GiB49±12.9%
●From the file◐Predictedwhat these mean

Speed is modeled, not measured: decode is memory-bandwidth bound, so tokens per second is bytes read per token against achievable bandwidth. Mixture-of-experts models carry a wider band because only the routed experts are read each step, and few have been measured publicly.

Questions people ask

What AI models can a GeForce RTX 5060 Ti run?
1400 of 2118 indexed open-weight models fit a GeForce RTX 5060 Ti at 4,096 context with f16 KV cache, the largest being DeepSeek-R1-Distill-Llama-8B-Abliterated at I1-IQ3_XXS. That covers text, vision-language, image, video and speech models.
How much usable memory does a GeForce RTX 5060 Ti actually have?
Its nameplate is 8 GB, but about 7.44 GiB is available to a model once driver and compositor overhead is accounted for.
Is a GeForce RTX 5060 Ti fast for local AI?
Its memory bandwidth is 448 GB/s, and that figure — not teraflops — is what governs token generation speed. Capacity decides what you can run; bandwidth decides how fast it runs.