What are the best models to run on NVIDIA RTX 5060 (8GB)?

Based on model rankings and 8 GB usable VRAM: Top Overall: Qwen3.5-9B (INT4) ~61 tok/s. Top Coding: K2 Horizon 7B (INT4) ~74 tok/s. Optimal Daily Driver: Granite 4.2 8B (INT4) ~67 tok/s. Maximum Model Size: Gemma 3 12B (INT4) ~43 tok/s.

How are VRAM, throughput, and latency calculated?

Calculations are determined using apxml.com modeling, combining LLM architecture analysis with GPU hardware physics. VRAM and Context Scaling: Memory requirements are determined from model architecture, including parameter count, layer configuration, and attention structure. The KV cache footprint scales deterministically based on context length, batch size, and cache precision. Throughput (TPS): Token generation is calculated in the memory-bandwidth-bound decode phase, measuring the time required to stream model weights and KV cache relative to hardware memory bandwidth. Latency (TTFT): Time to first token is calculated from the compute operations required during prompt prefill relative to hardware tensor compute capacity.

How accurate are these estimates?

Calculations are calibrated against empirical ground-truth benchmark runs across diverse architectures and hardware. Because the underlying mechanics are benchmark-calibrated, estimates remain accurate across different configurations without needing to run physical tests for every scenario. VRAM Accuracy: Memory estimates are deterministic and closely match actual allocation. Minor differences typically stay within 3% to 5% due to CUDA runtime overhead or memory allocator fragmentation. Speed and Latency Accuracy: Throughput and TTFT typically vary within 5% to 10% depending on the specific inference runtime, kernel implementations, and system power limits.

Best Models to Run on NVIDIA RTX 5060 (8GB) (October 2026)

Rank	Model	Parameters	Precision	TPS (Approx.)	TTFT
#113	Qwen3.5-9B	9B	INT4	61 tok/s	~1.3s
#118	K2 Horizon 7B	7B	INT4	74 tok/s	~1.0s
#128	NVIDIA Nemotron 3 Nano 30B-A3B	3.5B (30B active)	INT4 Q8	20 tok/s 11 tok/s	~4.1s ~4.5s
#132	Qwen3.5-4B	4B	INT4 Q8	111 tok/s 73 tok/s	~605 ms ~610 ms
#146	Granite 4.2 8B	8B	INT4	67 tok/s	~1.2s
#150	Gemma 4 E4B	8B	INT4	67 tok/s	~1.2s
#169	OLMo 3 7B Think	7B	INT4	50 tok/s	~1.1s
#177	Gemma 3 12B	12B	INT4	43 tok/s	~2.0s
#182	Ministral 3 8B	8B	INT4	43 tok/s	~1.4s
#183	Qwen3.5-2B	2B	INT4 Q8 FP16	166 tok/s 119 tok/s 73 tok/s	~324 ms ~326 ms ~331 ms
#184	MiniCPM5-2B	2B	INT4 Q8 FP16	166 tok/s 119 tok/s 73 tok/s	~324 ms ~326 ms ~331 ms
#188	Gemma 4 E2B	5.1B	INT4 Q8	94 tok/s 57 tok/s	~760 ms ~832 ms
#189	Qwen3-4B	4B	INT4 Q8	111 tok/s 73 tok/s	~605 ms ~610 ms
#190	DeepSeek-R1 8B	8B	INT4	69 tok/s	~1.2s
#193	Ministral 3 3B	3B	INT4 Q8	89 tok/s 69 tok/s	~467 ms ~471 ms
#197	Gemma 2 9B	9B	INT4	61 tok/s	~1.3s
#198	Qwen2.5-7B	7B	INT4	74 tok/s	~1.0s
#199	Qwen3-8B	8B	INT4	67 tok/s	~1.2s
#200	Gemma 3 4B	4B	INT4 Q8	111 tok/s 73 tok/s	~605 ms ~610 ms
#201	DeepSeek-R1 7B	7B	INT4 Q8	76 tok/s 42 tok/s	~1.0s ~1.2s
#202	Llama 3.1 8B	8B	INT4	67 tok/s	~1.2s
#204	Qwen3-1.7B	1.7B	INT4 Q8 FP16	180 tok/s 132 tok/s 83 tok/s	~278 ms ~280 ms ~284 ms
#206	Gemma 3 270M	0.27B	INT4 Q8 FP16	288 tok/s 265 tok/s 223 tok/s	~49 ms ~50 ms ~50 ms
#208	Llama 3 8B	8B	INT4	67 tok/s	~1.2s
#209	Qwen3-0.6B	0.6B	INT4 Q8 FP16	271 tok/s 227 tok/s 167 tok/s	~102 ms ~102 ms ~104 ms
#210	Phi-4-Mini	3.8B	INT4 Q8	115 tok/s 76 tok/s	~577 ms ~581 ms
#212	Ministral-8B-2410	8B	INT4	67 tok/s	~1.2s
#213	DeepSeek-R1 1.5B	1.5B	INT4 Q8 FP16	192 tok/s 143 tok/s 92 tok/s	~246 ms ~248 ms ~252 ms
#214	Llama 3.2 3B	3B	INT4 Q8 FP16	133 tok/s 90 tok/s 50 tok/s	~464 ms ~467 ms ~515 ms
#215	Qwen3.5-0.8B	0.8B	INT4 Q8 FP16	247 tok/s 200 tok/s 141 tok/s	~135 ms ~136 ms ~138 ms
#216	Phi-3-mini	3.8B	INT4 Q8	115 tok/s 76 tok/s	~577 ms ~581 ms
#217	Gemma 3n E2B IT	6B (2B active)	INT4	85 tok/s	~380 ms
#218	Mistral-7B-Instruct-v0.1	7.3B	INT4	72 tok/s	~1.1s
#219	Llama 3.2 1B	1B	INT4 Q8 FP16	227 tok/s 179 tok/s 122 tok/s	~170 ms ~171 ms ~174 ms
#220	Gemma 2 2B	2B	INT4 Q8 FP16	166 tok/s 119 tok/s 73 tok/s	~324 ms ~326 ms ~331 ms
#222	OLMo 3 7B Instruct	7B	INT4	50 tok/s	~1.1s
#224	Qwen2-7B	7B	INT4	74 tok/s	~1.0s
#225	Phi-3-small	7B	INT4	74 tok/s	~1.0s
#226	Gemma 3 1B	1B	INT4 Q8 FP16	227 tok/s 179 tok/s 122 tok/s	~170 ms ~171 ms ~174 ms
#227	Mistral-7B-Instruct-v0.2	7.3B	INT4	72 tok/s	~1.1s
#229	Gemma 1 7B	7B	INT4	50 tok/s	~1.1s
#230	Gemma 1 2B	2B	INT4 Q8 FP16	176 tok/s 124 tok/s 75 tok/s	~324 ms ~326 ms ~331 ms
#231	ChatGLM3-6B	6B	INT4	56 tok/s	~968 ms
#232	ChatGLM2-6B	6B	INT4	56 tok/s	~968 ms
#233	ChatGLM-6B	6B	INT4	56 tok/s	~968 ms
Unranked Models (Sorted by Speed)
-	Qwen2-0.5B	0.5B	INT4 Q8 FP16	284 tok/s 243 tok/s 184 tok/s	~85 ms ~86 ms ~87 ms
-	Qwen2.5-0.5B	0.5B	INT4 Q8 FP16	284 tok/s 243 tok/s 184 tok/s	~85 ms ~86 ms ~87 ms
-	ERNIE-4.5-0.3B	0.3B	INT4 Q8 FP16	279 tok/s 256 tok/s 213 tok/s	~54 ms ~54 ms ~55 ms
-	ERNIE-4.5-0.3B-Base	0.3B	INT4 Q8 FP16	279 tok/s 256 tok/s 213 tok/s	~54 ms ~54 ms ~55 ms
-	Falcon-1B	1B	INT4 Q8 FP16	238 tok/s 186 tok/s 125 tok/s	~170 ms ~171 ms ~174 ms
-	Falcon3-1B	1B	INT4 Q8 FP16	227 tok/s 179 tok/s 122 tok/s	~170 ms ~171 ms ~174 ms
-	Qwen2-1.5B	1.5B	INT4 Q8 FP16	192 tok/s 143 tok/s 92 tok/s	~246 ms ~248 ms ~252 ms
-	Qwen2.5-1.5B	1.5B	INT4 Q8 FP16	192 tok/s 143 tok/s 92 tok/s	~246 ms ~248 ms ~252 ms
-	Spark X2.5 1.7B	1.7B	INT4 Q8 FP16	180 tok/s 132 tok/s 83 tok/s	~278 ms ~280 ms ~284 ms
-	Hy-MT2 1.8B	1.8B	INT4 Q8 FP16	175 tok/s 128 tok/s 80 tok/s	~292 ms ~294 ms ~299 ms
-	Falcon-3B	3B	INT4 Q8 FP16	141 tok/s 94 tok/s 51 tok/s	~463 ms ~467 ms ~515 ms
-	CroissantLLM Base	1.3B	INT4 Q8 FP16	138 tok/s 117 tok/s 83 tok/s	~218 ms ~219 ms ~223 ms
-	Phi-1	1.3B	INT4 Q8 FP16	138 tok/s 117 tok/s 83 tok/s	~218 ms ~219 ms ~223 ms
-	Phi-1.5	1.3B	INT4 Q8 FP16	138 tok/s 117 tok/s 83 tok/s	~218 ms ~219 ms ~223 ms
-	DeepSeek-R1 3B	3B	INT4 Q8 FP16	135 tok/s 91 tok/s 50 tok/s	~464 ms ~467 ms ~515 ms
-	Falcon3-3B	3B	INT4 Q8 FP16	133 tok/s 90 tok/s 50 tok/s	~464 ms ~467 ms ~515 ms
-	Qwen2.5-3B	3B	INT4 Q8 FP16	133 tok/s 90 tok/s 50 tok/s	~464 ms ~467 ms ~515 ms
-	Spark X2.5 4B	4B	INT4 Q8	111 tok/s 73 tok/s	~605 ms ~610 ms
-	Phi-2	2.7B	INT4 Q8 FP16	94 tok/s 74 tok/s 44 tok/s	~425 ms ~428 ms ~510 ms
-	MaLLaM-3B	3B	INT4 Q8	89 tok/s 69 tok/s	~467 ms ~471 ms
-	SmolLM3 3B	3B	INT4 Q8	89 tok/s 69 tok/s	~467 ms ~471 ms
-	Falcon-7B	7B	INT4 Q8	78 tok/s 42 tok/s	~1.0s ~1.2s
-	Falcon3-7B	7B	INT4	74 tok/s	~1.0s
-	Hy-MT2 7B	7B	INT4	74 tok/s	~1.0s
-	Mistral-7B-v0.1	7.3B	INT4	72 tok/s	~1.1s
-	LensVLM 9B	9B	INT4	61 tok/s	~1.3s
-	MiMo V2.6 Distill Qwen 9B	9B	INT4	61 tok/s	~1.3s
-	ChatGLM3-6B-32K	6B	INT4	56 tok/s	~968 ms
-	Yi-6B	6B	INT4	56 tok/s	~968 ms
-	Falcon3-10B	10B	INT4	53 tok/s	~1.6s
-	Falcon2-11B	11B	INT4	51 tok/s	~1.7s
-	Hunyuan Lite	7B	INT4	50 tok/s	~1.1s
-	OLMo 3 7B Base	7B	INT4	50 tok/s	~1.1s
-	SEA-LION-7B	7.1B	INT4	50 tok/s	~1.1s
-	SEA-LION-7B-Instruct	7.1B	INT4	50 tok/s	~1.1s
-	Sahabat-AI-Llama3-8B-Instruct	8B	INT4	43 tok/s	~1.4s
-	Typhoon-2-8B	8B	INT4	43 tok/s	~1.4s

NVIDIA RTX 5060 (8GB)

VRAM Scaling by Context Length

Best Models to Run on NVIDIA RTX 5060 (8GB)

Inference Capacity

Fine-Tuning Capacity

Workload Recommendations

Frequently Asked Questions