Speech-to-text, TTS, and audio analysis
49
Models
6
Benchmarks
31
Providers
48
Open Weight
49
Deployable Now
48 self-host
49
Structured benchmark-backed rows
0
Signal-backed rows without full normalized tables
0
Benchmark-expected rows still catching up
| # | Model | Value | Quality | Params | SWE-Bench | Open ASR Mean WER | GAIA | MMMU | $/M In | Speed | Open |
|---|---|---|---|---|---|---|---|---|---|---|---|
18 | cohere-transcribe-03-2026 Cohere Open Weights |
| 52.7 |
| 52.7 |
| 2B |
| 42.4 |
| 94.6 |
| — |
| — |
Free Free |
| — |
17 | whisper-small.en OpenAI Open WeightsStructured | 45.1 | 45.1 | 242M | — | 91.4 | — | — | Free Free | — |
37 | Qwen3-ASR-0.6B-hf Qwen Open WeightsStructured | 45.0 | 45.0 | 782M | — | — | 44.2 | — | Free Free | — |
38 | Qwen3-TTS-12Hz-0.6B-Base Qwen Open WeightsStructured | 43.8 | 43.8 | 915M | — | — | 44.2 | — | Free Free | — |
39 | Qwen3-TTS-12Hz-1.7B-VoiceDesign Qwen Open WeightsStructured | 43.7 | 43.7 | 2B | — | — | 44.2 | — | Free Free | — |
42 | Qwen3-ForcedAligner-0.6B Qwen Open WeightsStructured | 43.1 | 43.1 | 918M | — | — | 44.2 | — | Free Free | — |
51 | canary-qwen-2.5b NVIDIA Open WeightsStructured | 35.8 | 35.8 | 3B | 40.2 | — | 4.7 | — | Free Free | — |
6 | GPT Audio OpenAI SubscriptionStructured | 26.3 | 17.1 | Undisclosed | 42.4 | — | — | — | $2.50/M Cheapest verified | — |
9 | vstr Alignment-Lab-AI Open WeightsStructured | — | — | 442M | 56.4 | — | — | — | Free Free | — |
10 | qwen3-asr-0.6b-GGUF cstr Open WeightsStructured | — | — | 600M | — | — | 44.2 | — | Free Free | — |
11 | qwen3-asr-1.7b-GGUF cstr Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
12 | qwen3-TTS-KSA vadimbelsky Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
13 | Qwen3-ASR-1.7B-nvfp4 vrfai Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
14 | Qwen3-ASR-1.7B-JA neosophie Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
15 | Qwen3-TTS-GGUF Serveurperso Open WeightsStructured | — | — | Unknown | — | — | 44.2 | — | Free Free | — |
16 | qwen3-tts-0.6b-base-GGUF cstr Open WeightsStructured | — | — | 600M | — | — | 44.2 | — | Free Free | — |
17 | qwen3-tts-1.7b-base-GGUF cstr Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
18 | Qwen3-ASR-0.6B-int8-QAD vrfai Open WeightsStructured | — | — | 600M | — | — | 44.2 | — | Free Free | — |
19 | qwen3-tts-emotion-tags SpragAI Open WeightsStructured | — | — | Unknown | — | — | 44.2 | — | Free Free | — |
20 | qwen3-asr-0.6b-onnx andrewleech Open WeightsStructured | — | — | 600M | — | — | 44.2 | — | Free Free | — |
21 | qwen3-asr-1.7b-onnx andrewleech Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
22 | Qwen3-ASR-1.7B-hf-beam4 bosonai Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
23 | qwen3-tts-coreml FluidInference Open WeightsStructured | — | — | Unknown | — | — | 44.2 | — | Free Free | — |
24 | Whisper-Large-V3-Turbo qualcomm Open WeightsStructured | — | — | Unknown | — | 92.2 | — | — | Free Free | — |
25 | qwen3.5-TTS-Emirati vadimbelsky Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
26 | Qwen3-ASR-0.6B-Albanian Kushtrim Open WeightsStructured | — | — | 782M | — | — | 44.2 | — | Free Free | — |
27 | whisper-large-v3-turbo metythorn Open WeightsStructured | — | — | 809M | — | 92.2 | — | — | Free Free | — |
28 | Qwen3-ASR-1.7B-MLX-4bit aufklarer Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
29 | granite-4.0-1b-speech ibm-granite Open WeightsStructured | — | — | 2B | 42.4 | — | — | — | Free Free | — |
30 | Qwen3-ASR-Enhanced-v0.1 mrfakename Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
31 | Qwen3-ASR-1.7B-Lyrics-Viet sunbv56 Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
32 | qwen3-tts-0.6b-indonesian wt5719001 Open WeightsStructured | — | — | 906M | — | — | 44.2 | — | Free Free | — |
33 | qwen3-tts-1.7b-voicedesign-GGUF cstr Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
34 | qwen3-tts-0-6b-base-mix-precision Arm Open WeightsStructured | — | — | 6B | — | — | 44.2 | — | Free Free | — |
35 | Qwen3-tts-0.6b-gguf-for-koboldcpp Jahaz Open WeightsStructured | — | — | 600M | — | — | 44.2 | — | Free Free | — |
36 | Qwen3-ASR-0.6B-Backpack-ASR backpack-run Open WeightsStructured | — | — | 938M | — | — | 44.2 | — | Free Free | — |
37 | qwen3-edgellm-jetson-artifacts harvestsu Open WeightsStructured | — | — | Unknown | — | — | 44.2 | — | Free Free | — |
38 | Qwen3-TTS-12Hz-0.6B-Base-bf16 mlx-community Open WeightsStructured | — | — | 600M | — | — | 44.2 | — | Free Free | — |
39 | Qwen3-TTS-12Hz-1.7B-Base-8bit mlx-community Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
40 | qwen3-tts-0-6b-custom-voice-mix-precision Arm Open WeightsStructured | — | — | 6B | — | — | 44.2 | — | Free Free | — |
41 | Qwen3-ASR-1.7B_Jordanian_Dialect_Arabic sarapd Open WeightsStructured | — | — | 7B | — | — | 44.2 | — | Free Free | — |
42 | whisper-large-v3-turbo-bulgarian-bulgaria sam8000 Open WeightsStructured | — | — | 809M | — | — | — | 54.0 | Free Free | — |
43 | Qwen3-TTS-12Hz-1.7B-CustomVoice-6bit mlx-community Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
44 | Qwen3-TTS-12Hz-1.7B-CustomVoice-8bit mlx-community Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
45 | Qwen3-TTS-12Hz-1.7B-CustomVoice-bf16 mlx-community Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
46 | Qwen3-TTS-12Hz-1.7B-VoiceDesign-8bit mlx-community Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
47 | Qwen3-TTS-12Hz-1.7B-VoiceDesign-bf16 mlx-community Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
48 | NOESIS-Qwen3-TTS-1.7B-Talker-Combo-LLM2-ASR2-BF16 AMAImedia Open WeightsStructured | — | — | 2B | — | — | 44.2 | — | Free Free | — |
49 | whisper-large-v3 ivrit-ai Open WeightsStructured | — | — | 2B | — | 92.6 | — | — | Free Free | — |