Hardware / NVIDIA /GeForce RTX 4060 Ti 16GB

NVIDIA · GPU

GeForce RTX 4060 Ti 16GB

Runs 86 models fully in memory, 20 more with CPU offload.

Memory

16 GB

Bandwidth

288 GB/s

Runs

86

01

Specifications

Memory

16 GB

Bandwidth

288 GB/s

FP16 compute

22.1 TFLOPS

TDP

165 W

MSRP

$499

Street price

—

Released

Jul 18, 2023

Architecture

Ada Lovelace

02

What it runs

Best quant per model — sorted by fit score.

ModelParamsBest quantContextSpeedVerdict

Llama 3.2 3B Instruct3.2BQ4_K_M131K119 tok/sestFits in memory Niagara 19m Batch.en20MQ4_K_M~262K19002 tok/sestFits in memory Niagara 38m Batch.en38MQ4_K_M~262K10001 tok/sestFits in memory ARK ASR 0.6B1.3BQ4_K_M~262K292 tok/sestFits in memory Audio8 ASR 0.1B0.3BQ8_0~262K723 tok/sestFits in memory Higgs Audio v3 STT2.7BQ5_K_M~66K120 tok/sestFits in memory Cohere Transcribe 03 20262.1BQ4_K_M~262K181 tok/sestFits in memory Distil Large v3.50.8BQ5_K_M~262K405 tok/sestFits in memory Lite Whisper Large v3 Acc1.4BQ8_0~262K155 tok/sestFits in memory Owsm CTC v3.1 1B1.1BQ8_0~262K197 tok/sestFits in memory Owsm CTC v3.2 ft 1B1BQ5_K_M~262K324 tok/sestFits in memory Owsm CTC v4 1B1BQ8_0~262K217 tok/sestFits in memory Hubert Large Ls960 ft0.3BQ8_0~262K678 tok/sestFits in memory MMS 1b All1BQ8_0~262K217 tok/sestFits in memory Wav2vec2 Large 960h Lv60 Self0.3BQ8_0~262K678 tok/sestFits in memory Granite 4.0 1b Speech2.3BQ5_K_M~131K141 tok/sestFits in memory Granite Speech 3.3 2b3BQ4_K_M~131K127 tok/sestFits in memory Granite Speech 4.1 2b2.3BQ5_K_M~131K141 tok/sestFits in memory Granite Speech 4.1 2b NAR2.3BQ4_K_M~131K165 tok/sestFits in memory STT 2.6b en2.6BQ5_K_M~262K125 tok/sestFits in memory Canary 180m Flash0.2BQ4_K_M~262K2111 tok/sestFits in memory Canary 1b1BQ4_K_M~262K380 tok/sestFits in memory Canary 1b Flash1BQ4_K_M~262K380 tok/sestFits in memory Canary 1b v21BQ8_0~262K217 tok/sestFits in memory Canary Qwen 2.5b2.6BQ4_K_M~262K146 tok/sestFits in memory Nemotron 3.5 ASR Streaming 0.6b0.6BQ4_K_M~262K594 tok/sestFits in memory Nemotron Speech Streaming en 0.6b0.6BQ8_0~262K361 tok/sestFits in memory Parakeet CTC 0.6b0.6BQ5_K_M~262K540 tok/sestFits in memory Parakeet CTC 1.1b1.1BQ5_K_M~262K295 tok/sestFits in memory Parakeet RNNT 0.6b0.6BQ8_0~262K361 tok/sestFits in memory Parakeet RNNT 1.1b1.1BQ4_K_M~262K346 tok/sestFits in memory Parakeet TDT CTC 110m0.1BQ8_0~262K1971 tok/sestFits in memory Parakeet TDT 0.6b v20.6BQ8_0~262K361 tok/sestFits in memory Parakeet TDT 0.6b v30.6BQ4_K_M~262K633 tok/sestFits in memory Parakeet TDT 1.1b1.1BQ5_K_M~262K295 tok/sestFits in memory Nyra Health CrisperWhisper1.6BQ4_K_M~262K238 tok/sestFits in memory Whisper Large v31.5BQ5_K_M~262K216 tok/sestFits in memory Whisper Large v3 Turbo0.8BQ8_0~262K271 tok/sestFits in memory MOSS Transcribe Diarize0.9BQ4_K_M~66K422 tok/sestFits in memory Qwen3 ASR 0.6B0.8BQ5_K_M~262K415 tok/sestFits in memory Qwen3 ASR 0.6B HF0.8BQ5_K_M~66K405 tok/sestFits in memory Qwen3 ASR 1.7B2BQ4_K_M~262K186 tok/sestFits in memory Qwen3 ASR 1.7B HF2BQ5_K_M~262K159 tok/sestFits in memory Zipformer cr CTC Transducer XL 290M0.3BQ8_0~262K748 tok/sestFits in memory Zipformer Transducer XL 290M0.3BQ4_K_M~262K1311 tok/sestFits in memory ASR Conformer Largescaleasr0.5BQ8_0~262K452 tok/sestFits in memory Moonshine Streaming Medium0.2BQ8_0~262K904 tok/sestFits in memory Moonshine Streaming Small0.1BQ5_K_M~262K3239 tok/sestFits in memory Moonshine Streaming Tiny30MQ5_K_M~262K10798 tok/sestFits in memory Moonshine Tiny30MQ8_0~262K7228 tok/sestFits in memory GLM ASR Nano 25122.3BQ5_K_M~131K141 tok/sestFits in memory Llama 3.2 1B Instruct1.2BQ5_K_M33K270 tok/sestFits in memory Ministral 3 3B 25123.8BQ4_K_M66K100 tok/sestFits in memory Granite 4.0 Micro3.2BQ4_K_M66K101 tok/sestFits in memory MOSS Transcribe Preview 2B2.4BQ5_K_M~16K135 tok/sestFits in memory ARK ASR 3B4.1BQ4_K_M~262K93 tok/sestFits in memory Gemma 3 4B4.3BQ4_K_M131K88 tok/sestFits in memory Voxtral Mini 4B Realtime 26024.4BQ4_K_M~66K86 tok/sestFits in memory Voxtral Mini 3B 25075BQ4_K_M~131K76 tok/sestFits in memory Hojo ASR V15.2BQ4_K_M~131K73 tok/sestFits in memory

A speed marked est is calculated from this card's memory bandwidth and the size of the weights it has to read per word. It is not a measurement.

Something wrong on this page? Tell us