Small Model Hub

3,378 efficient AI models across 7 domains · Charts · API

Domain

All (3,378)Vision (1,615)3D (771)Embedding (428)Multimodal (299)Code (125)Speech (91)Video (49)

Max Size≤10M ≤50M ≤100M ≤500M All

Max VRAM≤512MB ≤1GB ≤2GB ≤4GB ≤8GB All

Score

Raw Context (0-100)

Raw benchmark score — scale varies by task

Showing 68 models

#	Model	Domain	Task	Params ↑	GFLOPs	VRAM	Score	Speed	Arch	Source
1	nvidia/stt_en_conformer_ctc_small	Speech	Speech Recognition	10.0M	—	24 MB	——	6258 RTFx	Conformer-based/CTC	Open ASR
2	nvidia/stt_en_conformer_transducer_small	Speech	Speech Recognition	14.0M	—	34 MB	——	5557 RTFx	Conformer-based/TDT / RNN-T	Open ASR
3	abr-ai/niagara-19m-batch.en	Speech	Speech Recognition	20.0M	—	48 MB	——	3735 RTFx	—	Open ASR
4	usefulsensors/moonshine-streaming-tiny	Speech	Speech Recognition	30.0M	—	72 MB	——	4375 RTFx	Custom/Transformer	Open ASR
5	usefulsensors/moonshine-tiny	Speech	Speech Recognition	30.0M	—	72 MB	——	3733 RTFx	Custom/Transformer	Open ASR
6	abr-ai/niagara-38m-batch.en	Speech	Speech Recognition	38.0M	—	91 MB	——	4049 RTFx	—	Open ASR
7	usefulsensors/moonshine-base	Speech	Speech Recognition	60.0M	—	144 MB	——	2767 RTFx	Custom/Transformer	Open ASR
8	facebook/data2vec-audio-base-960h	Speech	Speech Recognition	90.0M	—	216 MB	——	2930 RTFx	Self-supervised/CTC	Open ASR
9	facebook/wav2vec2-base-960h	Speech	Speech Recognition	90.0M	—	216 MB	——	3366 RTFx	Self-supervised/CTC	Open ASR
10	nvidia/parakeet-tdt_ctc-110m	Speech	Speech Recognition	110.0M	—	264 MB	——	6119 RTFx	Conformer-based/"CTC	Open ASR
11	nvidia/stt_en_fastconformer_transducer_large	Speech	Speech Recognition	110.0M	—	264 MB	——	6033 RTFx	Conformer-based/TDT / RNN-T	Open ASR
12	nvidia/stt_en_conformer_ctc_large	Speech	Speech Recognition	120.0M	—	288 MB	——	5732 RTFx	Conformer-based/CTC	Open ASR
13	nvidia/stt_en_fastconformer_ctc_large	Speech	Speech Recognition	120.0M	—	288 MB	——	6027 RTFx	Conformer-based/CTC	Open ASR
14	usefulsensors/moonshine-streaming-small	Speech	Speech Recognition	120.0M	—	288 MB	——	3206 RTFx	Custom/Transformer	Open ASR
15	nvidia/canary-180m-flash	Speech	Speech Recognition	180.0M	—	432 MB	——	2484 RTFx	Conformer-based/Transformer	Open ASR
16	usefulsensors/moonshine-streaming-medium	Speech	Speech Recognition	240.0M	—	576 MB	——	2681 RTFx	Custom/Transformer	Open ASR
17	soundsgoodai/Zipformer-cr-ctc-transducer-XL-290M	Speech	Speech Recognition	290.0M	—	696 MB	——	160 RTFx	Zipformer/Pruned Stateless Transducer with k2 modified_beam_search	Open ASR
18	soundsgoodai/Zipformer-transducer-XL-290M	Speech	Speech Recognition	290.0M	—	696 MB	——	141 RTFx	Zipformer/RNN-T	Open ASR
19	facebook/wav2vec2-large-robust-ft-libri-960h	Speech	Speech Recognition	300.0M	—	720 MB	——	3058 RTFx	Self-supervised/CTC	Open ASR
20	facebook/data2vec-audio-large-960h	Speech	Speech Recognition	310.0M	—	744 MB	——	2374 RTFx	Self-supervised/CTC	Open ASR
21	facebook/hubert-large-ls960-ft	Speech	Speech Recognition	320.0M	—	768 MB	——	3024 RTFx	Self-supervised/CTC	Open ASR
22	facebook/wav2vec2-large-960h	Speech	Speech Recognition	320.0M	—	768 MB	——	3614 RTFx	Self-supervised/CTC	Open ASR
23	facebook/wav2vec2-large-960h-lv60-self	Speech	Speech Recognition	320.0M	—	768 MB	——	3010 RTFx	Self-supervised/CTC	Open ASR
24	speechbrain/asr-wav2vec2-librispeech	Speech	Speech Recognition	320.0M	—	768 MB	——	3628 RTFx	Self-supervised/CTC	Open ASR
25	AutoArk-AI/Audio8-ASR-0.1B	Speech	Speech Recognition	324.0M	—	778 MB	——	709 RTFx	Qwen3-ASR audio encoder/8-layer Qwen-style causal LM	Open ASR
26	speechbrain/asr-conformer-largescaleasr	Speech	Speech Recognition	480.0M	—	1.1 GB	——	72 RTFx	Conformer-based/"Transformer	Open ASR
27	facebook/wav2vec2-conformer-rope-large-960h-ft	Speech	Speech Recognition	600.0M	—	1.4 GB	——	1723 RTFx	"Self-supervised/Conformer-based"	Open ASR
28	nvidia/parakeet-ctc-0.6b	Speech	Speech Recognition	600.0M	—	1.4 GB	——	5884 RTFx	Conformer-based/CTC	Open ASR
29	nvidia/parakeet-rnnt-0.6b	Speech	Speech Recognition	600.0M	—	1.4 GB	——	5407 RTFx	Conformer-based/TDT / RNN-T	Open ASR
30	nvidia/parakeet-tdt-0.6b-v2	Speech	Speech Recognition	600.0M	—	1.4 GB	——	6038 RTFx	Conformer-based/TDT / RNN-T	Open ASR
31	nvidia/parakeet-tdt-0.6b-v3	Speech	Speech Recognition	600.0M	—	1.4 GB	——	6098 RTFx	Conformer-based/TDT / RNN-T	Open ASR
32	facebook/wav2vec2-conformer-rel-pos-large-960h-ft	Speech	Speech Recognition	620.0M	—	1.5 GB	——	1415 RTFx	"Self-supervised/Conformer-based"	Open ASR
33	nvidia/nemotron-speech-streaming-en-0.6b	Speech	Speech Recognition	620.0M	—	1.5 GB	——	1071 RTFx	FastConformer (cache-aware)/RNNT	Open ASR
34	nvidia/nemotron-3.5-asr-streaming-0.6b	Speech	Speech Recognition	640.0M	—	1.5 GB	——	1490 RTFx	FastConformer (cache-aware)/RNNT	Open ASR
35	efficient-speech/lite-whisper-large-v3-turbo-acc	Speech	Speech Recognition	700.0M	—	1.6 GB	——	327 RTFx	Whisper/Transformer	Open ASR
36	Qwen/Qwen3-ASR-0.6B	Speech	Speech Recognition	780.0M	—	1.8 GB	——	439 RTFx	Custom/Qwen3	Open ASR
37	Qwen/Qwen3-ASR-0.6B-hf	Speech	Speech Recognition	780.0M	—	1.8 GB	——	730 RTFx	Custom/Qwen3	Open ASR
38	distil-whisper/distil-large-v3.5	Speech	Speech Recognition	800.0M	—	1.9 GB	——	874 RTFx	Whisper/Transformer	Open ASR
39	openai/whisper-large-v3-turbo	Speech	Speech Recognition	800.0M	—	1.9 GB	——	783 RTFx	Whisper/Transformer	Open ASR
40	facebook/hubert-xlarge-ls960-ft	Speech	Speech Recognition	960.0M	—	2.3 GB	——	1980 RTFx	Self-supervised/CTC	Open ASR
41	facebook/mms-1b-all	Speech	Speech Recognition	960.0M	—	2.3 GB	——	1954 RTFx	Self-supervised/CTC	Open ASR
42	efficient-speech/lite-whisper-large-v3	Speech	Speech Recognition	1.0B	—	2.3 GB	——	212 RTFx	Whisper/Transformer	Open ASR
43	efficient-speech/lite-whisper-large-v3-acc	Speech	Speech Recognition	1.0B	—	2.3 GB	——	204 RTFx	Whisper/Transformer	Open ASR
44	efficient-speech/lite-whisper-large-v3-fast	Speech	Speech Recognition	1.0B	—	2.3 GB	——	212 RTFx	Whisper/Transformer	Open ASR
45	espnet/owsm_ctc_v3.2_ft_1B	Speech	Speech Recognition	1.0B	—	2.3 GB	——	692 RTFx	Whisper/CTC	Open ASR
46	espnet/owsm_ctc_v4_1B	Speech	Speech Recognition	1.0B	—	2.3 GB	——	765 RTFx	Whisper/CTC	Open ASR
47	nvidia/canary-1b	Speech	Speech Recognition	1.0B	—	2.3 GB	——	766 RTFx	Conformer-based/Transformer	Open ASR
48	nvidia/canary-1b-flash	Speech	Speech Recognition	1.0B	—	2.3 GB	——	2126 RTFx	Conformer-based/Transformer	Open ASR
49	nvidia/canary-1b-v2	Speech	Speech Recognition	1.0B	—	2.3 GB	——	1821 RTFx	Conformer-based/Transformer	Open ASR
50	espnet/owsm_ctc_v3.1_1B	Speech	Speech Recognition	1.1B	—	2.6 GB	——	816 RTFx	Whisper/CTC	Open ASR
51	nvidia/parakeet-ctc-1.1b	Speech	Speech Recognition	1.1B	—	2.6 GB	——	5015 RTFx	Conformer-based/CTC	Open ASR
52	nvidia/parakeet-rnnt-1.1b	Speech	Speech Recognition	1.1B	—	2.6 GB	——	4122 RTFx	Conformer-based/TDT / RNN-T	Open ASR
53	nvidia/parakeet-tdt-1.1b	Speech	Speech Recognition	1.1B	—	2.6 GB	——	4529 RTFx	Conformer-based/TDT / RNN-T	Open ASR
54	AutoArk-AI/ARK-ASR-0.6B	Speech	Speech Recognition	1.1B	—	2.7 GB	——	672 RTFx	Whisper/Qwen2	Open ASR
55	CohereLabs/cohere-transcribe-03-2026	Speech	Speech Recognition	2.0B	—	4.7 GB	——	916 RTFx	Conformer-based/Transformer	Open ASR
56	ibm-granite/granite-4.0-1b-speech	Speech	Speech Recognition	2.0B	—	4.7 GB	——	661 RTFx	Conformer-based/Granite	Open ASR
57	ibm-granite/granite-speech-4.1-2b	Speech	Speech Recognition	2.0B	—	4.7 GB	——	547 RTFx	Conformer-based/Granite	Open ASR
58	ibm-granite/granite-speech-4.1-2b-nar	Speech	Speech Recognition	2.0B	—	4.7 GB	——	2079 RTFx	Conformer-based/Hybrid CTC-LLM	Open ASR
59	nyrahealth/CrisperWhisper	Speech	Speech Recognition	2.0B	—	4.7 GB	——	33 RTFx	Whisper/Transformer	Open ASR
60	openai/whisper-large-v3	Speech	Speech Recognition	2.0B	—	4.7 GB	——	462 RTFx	Whisper/Transformer	Open ASR
61	zai-org/GLM-ASR-Nano-2512	Speech	Speech Recognition	2.0B	—	4.7 GB	——	333 RTFx	Whisper/Llama	Open ASR
62	Qwen/Qwen3-ASR-1.7B	Speech	Speech Recognition	2.0B	—	4.8 GB	——	394 RTFx	Custom/Qwen3	Open ASR
63	Qwen/Qwen3-ASR-1.7B-hf	Speech	Speech Recognition	2.0B	—	4.8 GB	——	796 RTFx	Custom/Qwen3	Open ASR
64	OpenMOSS-Team/MOSS-Transcribe-preview-2B	Speech	Speech Recognition	2.4B	—	5.7 GB	——	149 RTFx	Qwen3-Omni-MoE audio encoder/Qwen3	Open ASR
65	nvidia/canary-qwen-2.5b	Speech	Speech Recognition	2.5B	—	5.9 GB	——	861 RTFx	Conformer-based/Qwen2	Open ASR
66	kyutai/stt-2.6b-en	Speech	Speech Recognition	2.6B	—	6.1 GB	——	133 RTFx	Mimi tokenizer/Transformer	Open ASR
67	bosonai/higgs-audio-v3-stt	Speech	Speech Recognition	2.7B	—	6.3 GB	——	110 RTFx	Whisper/Qwen3	Open ASR
68	ibm-granite/granite-speech-3.3-2b	Speech	Speech Recognition	3.0B	—	7.0 GB	——	507 RTFx	Conformer-based/Granite	Open ASR