Pandrator 0.12.0 · 2026-10-16

Models and languages

Find speech, transcription and sound models by language and capability. Check the exact model, reference requirements and runtime before starting a project.

Listen to selected models

Search a model or family

Language Voice mode

51 model families

Model	Listed languages	Controls	Source and terms
Audio8 TTSaudio.cppOptional reference Exact variants (1) `audio8_tts_preview_0_6b_q8_0`q8_0	Documented languages German, English, Spanish, French, Italian +6 Japanese (ja), Korean (ko), Dutch (nl), Polish (pl), Yue Chinese (yue), Chinese (zh)	Sampling · Text chunking · Seed Generation parameters `max_tokenstext_chunk_sizetext_chunk_modetop_ptop_ktemperatureseed`	Apache-2.0 Model card and terms audio.cpp
Breeze TTSaudio.cppOptional reference · Voice design Exact variants (3) `breeze_tts_2_bf16`bf16 `breeze_tts_2_q4_0`q4_0 `breeze_tts_2_q8_0`q8_0	Documented languages English, Chinese	Delivery instructions Voice design Emotion in descriptions or directions Sampling · Guidance strength · Text chunking · Seed Generation parameters `text_chunk_sizetext_chunk_modemax_tokensguidance_scaletemperaturedepth_temperaturetop_ktop_pseed` Vocal events laugh, cough, clear_throat, sigh	BreezeBlue Research and Non-Commercial License 1.1 Model card and terms audio.cpp
Chatterboxaudio.cppReference voice Exact variants (3) `chatterbox_f16`f16 `chatterbox_q8_0`q8_0 `chatterbox_safetensors`native	Documented languages Arabic, Danish, German, Greek, English +14 Spanish (es), Finnish (fi), French (fr), Hindi (hi), Italian (it), Korean (ko), Malay (ms), Dutch (nl), Norwegian (no), Polish (pl), Portuguese (pt), Swedish (sv), Swahili (sw), Turkish (tr)	Expressiveness · Sampling · Guidance strength · Text chunking · Seed Generation parameters `exaggerationguidance_scaletemperaturerepetition_penaltymin_ptop_ps3gen_cfg_ratemax_tokensdo_samplegreedystop_on_eosseedtext_chunk_size`	MIT Model card and terms audio.cpp
Chatterbox Turboaudio.cppPreset voice Exact variants (2) `chatterbox_turbo_f16`f16 `chatterbox_turbo_q8_0`q8_0	Documented languages English	Sampling · Seed Generation parameters `seedtemperaturetop_prepetition_penalty` Vocal events laugh, chuckle, sigh, cough, gasp, clear_throat	MIT Model card and terms audio.cpp
Confucius4 TTSaudio.cppReference voice Exact variants (1) `confucius4_tts_orig`orig	Documented languages German, English, Spanish, French, Indonesian +9 Italian (it), Japanese (ja), Korean (ko), Malay (ms), Portuguese (pt), Russian (ru), Thai (th), Vietnamese (vi), Chinese (zh)	Sampling · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `temperaturetop_ptop_knum_beamsrepetition_penaltymax_tokensnum_inference_stepsguidance_scaletext_chunk_sizetext_chunk_modecross_fade_duration_secedge_fade_duration_secedge_pad_duration_secseed`	Apache-2.0 Model card and terms audio.cpp
CosyVoice3audio.cppReference voice Exact variants (2) `cosyvoice3_f32`f32 `cosyvoice3_q8_0`q8_0	Documented languages German, English, Spanish, French, Italian +5 Japanese (ja), Korean (ko), Russian (ru), Yue Chinese (yue), Chinese (zh)	Delivery instructions Emotion in descriptions or directions Sampling · Synthesis steps · Text chunking · Seed Generation parameters `text_chunk_sizetext_chunk_modemax_tokensmin_tokenstop_knum_inference_stepsseed`	Apache-2.0 Model card and terms audio.cpp
dots.ttsaudio.cppReference voice Exact variants (4) `dots_tts_mf_bf16`bf16 Reference voice `dots_tts_mf_q8_0`q8_0 Reference voice `dots_tts_soar_bf16`bf16 Reference voice Variant source `dots_tts_soar_q8_0`q8_0 Reference voice Variant source	Partial list Arabic, Czech, German, Greek, English +19 Spanish (es), Finnish (fi), French (fr), Hindi (hi), Indonesian (id), Italian (it), Japanese (ja), Korean (ko), Dutch (nl), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Thai (th), Turkish (tr), Ukrainian (uk), Vietnamese (vi), Yue Chinese (yue), Chinese (zh)	Controls vary by variant Compare variants DotTTS MeanFlow dots_tts_mf_bf16, dots_tts_mf_q8_0 Synthesis steps · Text chunking · Seed Generation parameters `reference_duration_secnum_inference_stepsspeaker_scalesampler_modemax_tokenstext_chunk_sizetext_chunk_modeseed` DotTTS SOAR dots_tts_soar_bf16, dots_tts_soar_q8_0 Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `reference_duration_secnum_inference_stepsguidance_scalespeaker_scalesampler_modemax_tokenstext_chunk_sizetext_chunk_modeseed`	Apache-2.0 Model card and terms audio.cpp
DramaBoxaudio.cppOptional reference Exact variants (1) `dramabox_q8_0`q8_0	Documented languages English	Duration · Guidance strength · Synthesis steps · Seed Generation parameters `negative_promptguidance_rescaleduration_secnum_inference_stepsguidance_scalespatio_temporal_guidance_scaleduration_scalereference_duration_secaudio_chunk_threshold_secaudio_chunk_duration_seccross_fade_duration_secseed`	LTX-2 Community License Model card and terms audio.cpp
Echo-TTSaudio.cppReference voice Exact variants (2) `echo_tts_f16`f16 `echo_tts_q8_0`q8_0	Documented languages English	Guidance strength · Synthesis steps · Seed Generation parameters `text_guidance_scalespeaker_guidance_scalenum_inference_stepstruncation_factorspeaker_kv_scaleseedreference_duration_secmax_duration_secguidance_interval`	CC-BY-NC-SA-4.0 Model card and terms audio.cpp
FireRedAudioaudio.cppReference voice Exact variants (2) `firered_audio_orig`orig `firered_audio_q8_0`q8_0	Documented languages English, Chinese	Sampling · Guidance strength · Synthesis steps · Seed Generation parameters `num_inference_stepsguidance_scalemax_new_audio_stepsmin_new_audio_stepsmax_new_text_tokensmax_new_tokensenable_thinkingtop_ktop_ptemperatureseed`	Apache-2.0 Model card and terms audio.cpp
FireRedTTS3audio.cppReference voice · Optional reference · Voice design Exact variants (4) `fireredtts3_base_orig`orig Reference voice `fireredtts3_base_q8_0`q8_0 Reference voice `fireredtts3_instruct_orig`orig Optional reference · Voice design `fireredtts3_instruct_q8_0`q8_0 Optional reference · Voice design	Documented languages Arabic, Czech, German, Greek, English +19 Spanish (es), Finnish (fi), French (fr), Hindi (hi), Indonesian (id), Italian (it), Japanese (ja), Korean (ko), Dutch (nl), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Thai (th), Turkish (tr), Ukrainian (uk), Vietnamese (vi), Yue Chinese (yue), Chinese (zh)	Controls vary by variant Compare variants FireRedTTS3 Base Orig GGUF / FireRedTTS3 Base fireredtts3_base_orig, fireredtts3_base_q8_0 Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `seednum_inference_stepsguidance_scalestop_thresholdtext_chunk_sizetext_chunk_mode` FireRedTTS3 Instruct Orig GGUF / FireRedTTS3 Instruct fireredtts3_instruct_orig, fireredtts3_instruct_q8_0 Voice description Voice design Emotion in descriptions or directions Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `seednum_inference_stepsguidance_scalestop_thresholdtext_chunk_sizetext_chunk_mode`	Apache-2.0 Model card and terms audio.cpp
Fish Audio S2 Proaudio.cppOptional reference Exact variants (2) `fish_audio_s2_pro_bf16`bf16 `fish_audio_s2_pro_q8_0`q8_0	Source claim Afrikaans, Amharic, Arabic, Assamese, Azerbaijani +78 Belarusian (be), Bulgarian (bg), Bangla (bn), Tibetan (bo), Breton (br), Bosnian (bs), Catalan (ca), Czech (cs), Welsh (cy), Danish (da), German (de), Greek (el), English (en), Spanish (es), Estonian (et), Basque (eu), Persian (fa), Finnish (fi), Faroese (fo), French (fr), Galician (gl), Gujarati (gu), Hebrew (he), Hindi (hi), Croatian (hr), Haitian Creole (ht), Hungarian (hu), Armenian (hy), Indonesian (id), Icelandic (is), Italian (it), Japanese (ja), Javanese (jv), Georgian (ka), Kazakh (kk), Khmer (km), Kannada (kn), Korean (ko), Latin (la), Lithuanian (lt), Latvian (lv), Māori (mi), Malayalam (ml), Mongolian (mn), Marathi (mr), Malay (ms), Burmese (my), Nepali (ne), Dutch (nl), Norwegian Nynorsk (nn), Norwegian (no), Punjabi (pa), Polish (pl), Pashto (ps), Portuguese (pt), Romanian (ro), Russian (ru), Sanskrit (sa), Sindhi (sd), Sinhala (si), Slovak (sk), Slovenian (sl), Shona (sn), Albanian (sq), Serbian (sr), Swedish (sv), Swahili (sw), Tamil (ta), Telugu (te), Thai (th), Filipino (tl), Turkish (tr), Ukrainian (uk), Urdu (ur), Vietnamese (vi), Yiddish (yi), Yoruba (yo), Chinese (zh)	Delivery instructions Emotion in descriptions or directions Sampling · Text chunking · Seed Generation parameters `seedmax_tokenstop_ptop_ktemperaturetext_chunk_sizetext_chunk_mode` Vocal events pause, laugh, chuckle, sigh, inhale, exhale, clear_throat	Fish Audio Research License Model card and terms audio.cpp
GLM-TTSaudio.cppReference voice Exact variants (1) `glm_tts_q8_0`q8_0	Documented languages English, Chinese	Sampling · Synthesis steps · Seed Generation parameters `max_tokenstemperaturetop_ktop_pseednum_inference_stepsflow_guidance_scalehift_prior_noise_count`	MIT Model card and terms audio.cpp
Habibi TTS (F5)audio.cppReference voice Exact variants (8) `habibi_alg`orig `habibi_egy`orig `habibi_irq`orig `habibi_mar`orig `habibi_msa`orig `habibi_sau`orig `habibi_uae`orig `habibi_unified`orig	Documented languages Arabic, English	Speed · Dialect · Guidance strength · Synthesis steps · Seed Generation parameters `dialectsway_sampling_coefspeedseednum_inference_stepsguidance_scalestrip_diacritics`	Terms by variant Apache-2.0 habibi_alg, habibi_egy, habibi_irq, habibi_mar, habibi_msa Model terms CC-BY-NC-SA-4.0 habibi_sau, habibi_uae, habibi_unified Model terms Model card and terms audio.cpp
Higgs Audioaudio.cppReference voice Exact variants (2) `higgs_audio_tts_4b_bf16`bf16 `higgs_audio_tts_4b_q8_0`q8_0	Source claim Afrikaans, Arabic, Assamese, Asturian, Azerbaijani +97 Bashkir (ba), Belarusian (be), Bulgarian (bg), Bangla (bn), Bosnian (bs), Catalan (ca), Cebuano (ceb), Central Kurdish (ckb), Czech (cs), Welsh (cy), Danish (da), German (de), Greek (el), English (en), Esperanto (eo), Spanish (es), Estonian (et), Basque (eu), Persian (fa), Finnish (fi), French (fr), Irish (ga), Galician (gl), Gujarati (gu), Hausa (ha), Hebrew (he), Hindi (hi), Croatian (hr), Haitian Creole (ht), Hungarian (hu), Armenian (hy), Indonesian (id), Icelandic (is), Italian (it), Japanese (ja), Javanese (jv), Georgian (ka), Kabyle (kab), Kamba (Kenya) (kam), Kabuverdianu (kea), Kazakh (kk), Kannada (kn), Korean (ko), Kyrgyz (ky), Latin (la), Luxembourgish (lb), Ganda (lg), Lingala (ln), Lithuanian (lt), Luo (Kenya and Tanzania) (luo), Latvian (lv), Eastern Mari (mhr), Māori (mi), Macedonian (mk), Malayalam (ml), Mongolian (mn), Marathi (mr), Malay (ms), Maltese (mt), Nepali (ne), Dutch (nl), Norwegian (no), Pedi (nso), Nyanja (ny), Occitan (oc), Oromo (om), Punjabi (pa), Polish (pl), Pashto (ps), Portuguese (pt), Romanian (ro), Russian (ru), Kinyarwanda (rw), Sindhi (sd), Slovak (sk), Slovenian (sl), Shona (sn), Somali (so), Albanian (sq), Serbian (sr), Swedish (sv), Swahili (sw), Tamil (ta), Telugu (te), Tajik (tg), Thai (th), Filipino (tl), Turkish (tr), Uyghur (ug), Ukrainian (uk), Umbundu (umb), Urdu (ur), Uzbek (uz), Vietnamese (vi), Xhosa (xh), Chinese (zh), Zulu (zu)	Sampling · Text chunking · Seed Generation parameters `max_tokenstemperaturetop_ptop_kseedtext_chunk_sizetext_chunk_mode`	Research / Non-Commercial + Creator Use Grant Model card and terms audio.cpp
IndexTTSaudio.cppReference voice Exact variants (7) `index_tts2_5_f16`f16 Arabic, English, Spanish, Japanese, Chinese Reference voice `index_tts2_5_orig`orig Arabic, English, Spanish, Japanese, Chinese Reference voice `index_tts2_5_q8_0`q8_0 Arabic, English, Spanish, Japanese, Chinese Reference voice `index_tts2_f16`f16 English, Chinese Reference voice Variant source `index_tts2_orig`orig English, Chinese Reference voice Variant source `index_tts2_q8_0`q8_0 English, Chinese Reference voice Variant source `index_tts2_safetensors`native English, Chinese Reference voice Variant source	By variant Arabic, English, Spanish, Japanese, Chinese	Emotion choices Emotion · Duration · Sampling · Text chunking · Seed Generation parameters `emotion_alphaemotion_vectoruse_emotion_textemotion_textuse_random_emotioninterval_silence_msdo_sampletop_ptop_ktemperaturelength_penaltynum_beamsrepetition_penaltymax_tokensseedtext_chunk_sizetext_chunk_modeduration_factor`	Terms by variant Bilibili Model License (IndexTTS 2.5) index_tts2_5_f16, index_tts2_5_orig, index_tts2_5_q8_0 Model terms Bilibili Model License (IndexTTS 2) index_tts2_f16, index_tts2_orig, index_tts2_q8_0, index_tts2_safetensors Model terms Model card and terms audio.cpp
Inflect Micro v2audio.cppPreset voice Exact variants (1) `inflect_micro_v2_orig`orig	Documented languages English	Speed · Text chunking · Seed Generation parameters `speaking_ratevariationseedtext_chunk_modetext_chunk_size`	Apache-2.0 Model card and terms audio.cpp
Irodori TTSaudio.cppOptional reference · Voice design Exact variants (6) `irodori_tts_500m_v3_f16`f16 Optional reference `irodori_tts_500m_v3_q8_0`q8_0 Optional reference `irodori_tts_600m_v3_voicedesign_f16`f16 Optional reference · Voice design Variant source `irodori_tts_600m_v3_voicedesign_q8_0`q8_0 Optional reference · Voice design Variant source `irodori_tts_v4_small_f16`f16 Optional reference · Voice design Variant source `irodori_tts_v4_small_q8_0`q8_0 Optional reference · Voice design Variant source	Documented languages Japanese	Controls vary by variant Compare variants Irodori-TTS 500M v3 irodori_tts_500m_v3_f16, irodori_tts_500m_v3_q8_0 Duration · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `num_inference_stepsduration_secduration_scaletext_chunk_modetext_chunk_sizemin_duration_secmax_duration_sectext_guidance_scalespeaker_guidance_scalecaption_guidance_scaleguidance_scaleguidance_modeguidance_min_tguidance_max_tseedtrim_tail` Irodori-TTS 600M v3 VoiceDesign / Irodori-TTS v4.1 Small irodori_tts_600m_v3_voicedesign_f16, irodori_tts_600m_v3_voicedesign_q8_0, irodori_tts_v4_small_f16, irodori_tts_v4_small_q8_0 Voice description Voice design Emotion in descriptions or directions Duration · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `num_inference_stepsduration_secduration_scaletext_chunk_modetext_chunk_sizemin_duration_secmax_duration_sectext_guidance_scalespeaker_guidance_scalecaption_guidance_scaleguidance_scaleguidance_modeguidance_min_tguidance_max_tseedtrim_tail`	MIT Model card and terms audio.cpp
Kitten TTS 2audio.cppHybrid mode Exact variants (1) `kitten_tts2_q8_0`q8_0	Documented languages Arabic, German, English, Spanish, French +5 Hindi (hi), Italian (it), Portuguese (pt), Russian (ru), Chinese (zh)	Sampling · Text chunking · Seed Generation parameters `presettemperaturetop_ktop_pmin_prepetition_penaltymax_tokensseedtext_chunk_sizetext_chunk_mode`	Stellon Labs Community License Model card and terms audio.cpp
Kitten TTS Miniaudio.cppPreset voice Exact variants (1) `kitten_tts_mini_0_8_orig`orig	Documented languages English	Speed · Text chunking · Seed Generation parameters `speedseedtext_chunk_modetext_chunk_size`	Apache-2.0 Model card and terms audio.cpp
Kokoroaudio.cppPreset voice Exact variants (2) `kokoro_82m_bf16`bf16 `kokoro_82m_q8_0`q8_0	Documented languages British English, American English, Spanish, French (France), Hindi +4 Italian (it), Japanese (ja), Brazilian Portuguese (pt-br), Chinese (zh) Japanese needs a GGUF with UniDic data and a MeCab library. The standard small package omits that dictionary. Package setup	Speed · Text chunking · Seed Generation parameters `seedspeedtext_chunk_sizereturn_timestamps`	Apache-2.0 Model card and terms audio.cpp
KugelAudioaudio.cppPreset voice Exact variants (3) `kugelaudio_0_open_bf16`bf16 `kugelaudio_0_open_q4_k`q4_k `kugelaudio_0_open_q8_0`q8_0	Documented languages Bulgarian, Czech, Danish, German, Greek +18 English (en), Spanish (es), Finnish (fi), French (fr), Croatian (hr), Hungarian (hu), Italian (it), Dutch (nl), Norwegian (no), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Slovak (sk), Serbian (sr), Swedish (sv), Turkish (tr), Ukrainian (uk)	Sampling · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `num_inference_stepsguidance_scalemax_tokensdo_sampletemperatureseedtext_chunk_sizetext_chunk_mode`	MIT Model card and terms audio.cpp
LFM2.5 Audioaudio.cppPreset voice Exact variants (7) `lfm2_audio_1_5b_f16`f16 English Preset voice `lfm2_audio_1_5b_jp_f16`f16 Japanese Preset voice Variant source `lfm2_audio_1_5b_jp_f32`f32 Japanese Preset voice Variant source `lfm2_audio_1_5b_jp_q4_0`q4_0 Japanese Preset voice Variant source `lfm2_audio_1_5b_jp_q8_0`q8_0 Japanese Preset voice Variant source `lfm2_audio_1_5b_q4_0`q4_0 English Preset voice `lfm2_audio_1_5b_q8_0`q8_0 English Preset voice	By variant English, Japanese	Sampling · Text chunking · Seed Generation parameters `max_tokenstemperaturetop_kseedtext_chunk_modetext_chunk_size`	LFM Open License v1.0 Model card and terms audio.cpp
Magpie TTSaudio.cppPreset voice Exact variants (2) `magpie_tts_orig`orig `magpie_tts_q8_0`q8_0	Partial list Arabic (United Arab Emirates), Arabic (Saudi Arabia), German, English, Spanish +7 French (fr), Hindi (hi), Italian (it), Korean (ko), Brazilian Portuguese (pt-br), Vietnamese (vi), Chinese (zh)	Sampling · Guidance strength · Text chunking · Seed Generation parameters `seedtemperaturetop_kguidance_scalemax_tokenstext_chunk_sizetext_chunk_mode`	NVIDIA Open Model License Model card and terms audio.cpp
Maya1audio.cppVoice design Exact variants (2) `maya1_orig`orig `maya1_q8_0`q8_0	Documented languages English	Delivery instructions Voice design Emotion in descriptions or directions Sampling · Text chunking · Seed Generation parameters `max_tokensmin_tokenstemperaturetop_prepetition_penaltyseedtext_chunk_sizetext_chunk_mode` Vocal events laugh, sigh, chuckle, gasp, exhale	Apache-2.0 Model card and terms audio.cpp
MioTTSaudio.cppReference voice Exact variants (3) `miotts_1_7b_bf16`bf16 `miotts_1_7b_orig`orig `miotts_1_7b_q8_0`q8_0	Documented languages English, Japanese	Sampling · Seed Generation parameters `max_tokenstop_ktop_ptemperaturerepetition_penaltypresence_penaltyfrequency_penaltydo_sampleseedbest_of_nbest_of_n_enabledbest_of_n_language`	Apache-2.0 Model card and terms audio.cpp
MOSS VoiceGenaudio.cppVoice design Exact variants (1) `moss_voicegen_bf16_codec_f16_decode`bf16	Documented languages English, Chinese	Voice description Voice design Emotion in descriptions or directions Text chunking · Seed Generation parameters `seedtext_chunk_sizetext_chunk_mode`	Apache-2.0 Model card and terms audio.cpp
MOSS-TTS Localaudio.cppOptional reference Exact variants (2) `moss_tts_local_v1_5_bf16`bf16 `moss_tts_local_v1_5_q8_0`q8_0	Documented languages Arabic, Czech, Danish, German, Greek +26 English (en), Spanish (es), Persian (fa), Finnish (fi), French (fr), Hebrew (he), Hindi (hi), Hungarian (hu), Italian (it), Japanese (ja), Korean (ko), Macedonian (mk), Malay (ms), Dutch (nl), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Swedish (sv), Swahili (sw), Thai (th), Filipino (tl), Turkish (tr), Vietnamese (vi), Yue Chinese (yue), Chinese (zh)	Sampling · Text chunking · Seed Generation parameters `max_tokensdo_sampletemperaturetop_ptop_krepetition_penaltytext_temperaturetext_top_ptext_top_kseedtext_chunk_sizetext_chunk_mode`	Apache-2.0 Model card and terms audio.cpp
MOSS-TTS Nanoaudio.cppOptional reference Exact variants (2) `moss_tts_nano_100m_bf16`bf16 `moss_tts_nano_100m_q8_0`q8_0	Documented languages Arabic, Czech, Danish, German, Greek +14 English (en), Spanish (es), Persian (fa), French (fr), Hungarian (hu), Italian (it), Japanese (ja), Korean (ko), Polish (pl), Portuguese (pt), Russian (ru), Swedish (sv), Turkish (tr), Chinese (zh)	Sampling · Text chunking · Seed Generation parameters `max_tokensactive_codebooksdo_sampletemperaturetop_ptop_krepetition_penaltytext_temperaturetext_top_ptext_top_kseedtext_chunk_sizetext_chunk_mode`	Apache-2.0 Model card and terms audio.cpp
MOSS-TTS v1.5audio.cppOptional reference Exact variants (3) `moss_tts_v15_bf16_codec_f16`bf16 `moss_tts_v15_q4_k_codec_f16`q4_k `moss_tts_v15_q8_0_codec_f16`q8_0	Documented languages English, Chinese	Voice description Emotion in descriptions or directions Sampling · Text chunking · Seed Generation parameters `tokensseedtemperaturetop_ptop_krepetition_penaltytext_chunk_sizetext_chunk_mode`	Apache-2.0 Model card and terms audio.cpp
MOSS-TTSDaudio.cppOptional reference Exact variants (3) `moss_ttsd_bf16_codec_f16`bf16 `moss_ttsd_q4_k_codec_f16`q4_k `moss_ttsd_q8_0_codec_f16`q8_0	Documented languages English, Chinese	Delivery instructions Sampling · Seed Generation parameters `seedtemperaturetop_ptop_krepetition_penaltymax_frames`	Apache-2.0 Model card and terms audio.cpp
NeuTTS 2Eaudio.cppPreset voice Exact variants (1) `neutts_2e_orig`orig	Documented languages English	Emotion choices Emotion · Sampling · Text chunking · Seed Generation parameters `emotionmax_tokensmin_tokenstemperaturetop_kseedtext_chunk_modetext_chunk_size`	Terms not verified Model card and terms audio.cpp
OmniVoiceaudio.cppOptional reference · Voice design Exact variants (4) `omnivoice_bf16`bf16 `omnivoice_f16`f16 `omnivoice_q8_0`q8_0 `omnivoice_safetensors`native	Documented languages Arbëreshë Albanian, Afade, Algerian Saharan Arabic, Abkhazian, Bankon +641 Abua (abn), Abron (abr), Ambonese Malay (abs), Baharna Arabic (abv), Mesopotamian Arabic (acm), Hijazi Arabic (acw), Omani Arabic (acx), Dhofari Arabic (adf), Amdo Tibetan (adx), Adyghe (ady), Tunisian Arabic (aeb), Saidi Arabic (aec), Afrikaans (af), Gulf Arabic (afb), Eloyi (afo), Igo (ahl), Ashe (ahs), Aja (Benin) (ajg), Judeo-Moroccan Arabic (aju), Alago (ala), Gheg Albanian (aln), Larike-Wakasihu (alo), Amharic (am), Guerrero Amuzgo (amu), Aragonese (an), Ngas (anc), Goemai (ank), Angika (anp), Anaang (anw), Ömie (aom), Levantine Arabic (apc), Sudanese Arabic (apd), Standard Arabic (arb), Algerian Arabic (arq), Najdi Arabic (ars), Moroccan Arabic (ary), Egyptian Arabic (arz), Assamese (as), Asturian (ast), Eastern Egyptian Bedawi Arabic (avl), Awak (awo), Libyan Arabic (ayl), North Mesopotamian Arabic (ayp), Azerbaijani (az), Bashkir (ba), Tuki (bag), Basa (Cameroon) (bas), Bamun (bax), Baatonum (bba), Ghomálá' (bbj), Bats (bbl), Kulung (Nigeria) (bbu), Bamenyam (bce), Baoulé (bci), Kohumono (bcs), Bacama (bcy), Bayot (bda), Bade (bde), Buduma (bdm), Belarusian (be), Bebele (beb), Betawi (bew), Bafut (bfd), Balti (bft), Bulgarian (bg), Eastern Balochi (bgp), Bhili (bhb), Bukharic (bhh), Bhojpuri (bho), Bima (bhp), Bara Malagasy (bhr), Kanauji (bjj), Barok (bjk), Banjar (bjn), Balanta-Ganja (bjt), Bakoko (bkh), Kom (Cameroon) (bkm), Bokyi (bky), Northern Betsimisaraka Malagasy (bmm), Bomu (bmq), Bangla (bn), Batanga (bnm), Bunun (bnn), Bundeli (bns), Tibetan (bo), Bondei (bou), Bago-Kusuntu (bqg), Breton (br), Braj (bra), Brahui (brh), Mokpwe (bri), Bodo (India) (brx), Bosnian (bs), Kati (bsh), Bangwinji (bsj), Burushaski (bsk), Batak Mandailing (btm), Bateri (btv), Buginese (bug), Bulu (Cameroon) (bum), Terei (buo), Boghom (bux), Bura-Pabir (bwr), Bilur (bxf), Ubaghara (byc), Burak (bys), Medumba (byv), Qaqet (byx), Southern Betsimisaraka Malagasy (bzc), Basa (Nigeria) (bzw), Catalan (ca), Samba Daka (ccg), Cebuano (ceb), Cen (cen), Dijim-Bwilim (cfa), Chiga (cgg), Quiotepec Chinantec (chq), Chokwe (cjk), Central Kurdish (ckb), Cibak (ckl), Kairak (ckr), Cakfem-Mushere (cky), Hakha Chin (cnh), South Ucayali Ashéninka (cpy), Czech (cs), Tepinapa Chinantec (cte), Tlacoatzintepec Chinantec (ctl), Teutila Cuicatec (cut), Tepeuxila Cuicatec (cux), Chuvash (cv), Welsh (cy), Danish (da), Dagbani (dag), Dargwa (dar), Taita (dav), Dadiya (dbd), Deccan (dcc), German (de), Degema (deg), Dghwede (dgh), Dogri (individual language) (dgo), Zarma (dje), Domaaki (dmk), Dameli (dml), Rukai (dru), Dotyali (dty), Duala (dua), Divehi (dv), Dyula (dyu), Dazaga (dzg), Ebrié (ebr), Embu (ebu), Eggon (ego), Askopan (eiv), Koti (eko), Yace (ekr), Greek (el), Eleme (elm), English (en), Esperanto (eo), Spanish (es), Central Yupik (esu), Estonian (et), Eton (Cameroon) (eto), Yekhee (ets), Ejagham (etu), Basque (eu), Ewondo (ewo), Extremaduran (ext), Keiyo (eyo), Persian (fa), Fang (Equatorial Guinea) (fan), Fanti (fat), Fula (ff), Maasina Fulfulde (ffm), Finnish (fi), Nobiin (fia), Filipino (fil), Fipa (fip), Kirya-Konzəl (fkk), Fe'fe' (fmp), French (fr), Adamawa Fulfulde (fub), Pulaar (fuc), Borgu Fulfulde (fue), Pular (fuf), Western Niger Fulfulde (fuh), Bagirmi Fulfulde (fui), Central-Eastern Niger Fulfulde (fuq), Nigerian Fulfulde (fuv), Western Frisian (fy), Irish (ga), Garhwali (gbm), Gbagyi (gbr), Gbari (gby), Mali (gcc), Guduf-Gava (gdf), Gen (gej), Geser-Gorom (ges), Gurgula (ggg), Gidar (gid), Goaria (gig), South Giziga (giz), Kachi Koli (gjk), Gujari (gju), Galician (gl), Glavda (glw), Guarani (gn), Gola (gol), Goan Konkani (gom), Gusilay (gsl), Gujarati (gu), Eastern Bolivian Guaraní (gui), Farefare (gur), Gusii (guz), Manx (gv), Gawri (gwc), Gweno (gwe), Gawar-Bati (gwt), Northwest Gbaya (gya), Geji (gyz), Hausa (ha), Hahon (hah), Hakö (hao), Hawaiian (haw), Hazaragi (haz), Huba (hbb), Hebrew (he), Hemba (hem), Hindi (hi), Lamang (hia), Hunjara-Kaina Ke (hkk), Halia (hla), Northern Hindko (hno), Hadothi (hoj), Croatian (hr), Upper Sorbian (hsb), Haitian Creole (ht), Hungarian (hu), San Francisco Del Mar Huave (hue), Hula (hul), Nüpode Huitoto (hux), Hwana (hwo), Armenian (hy), Herero (hz), Interlingua (ia), Ibibio (ibb), Indonesian (id), Idakho-Isukha-Tiriki (ida), Idoma (idu), Igbo (ig), Izon (ijc), Kalabari (ijn), Inupiaq (ik), Ikwere (ikw), Icelandic (is), Esan (ish), Isoko (iso), Italian (it), Isekiri (its), Ito (itw), Itzá (itz), Japanese (ja), Yalahatan (jal), Jambi Malay (jax), Ngomba (jgo), Western Juxtlahuaca Mixtec (jmx), Jaunsari (jns), Jaqaru (jqr), Wapan (juk), Jiba (juo), Javanese (jv), Georgian (ka), Kabyle (kab), Karekare (kai), Jju (kaj), Kamba (Kenya) (kam), Kabardian (kbd), Kanembu (kbl), Abadi (kbt), Kamo (kcq), Tem (kdh), Kabuverdianu (kea), Akebu (keu), Kota (India) (kfe), Kinnauri (kfk), Korwa (kfp), Khams Tibetan (khg), Khowar (khw), Kuanyama (kj), Coastal Konjo (kjc), Highland Konjo (kjk), Kazakh (kk), Kalenjin (kln), Kalasha (kls), Khmer (km), Northern Kurdish (kmr), Koma (kmy), Kannada (kn), Dera (Nigeria) (kna), Konkani (individual language) (knn), Korean (ko), Kol (Papua New Guinea) (kol), Konzo (koo), Ikposo (kpo), Eastern Krahn (kqo), Kashmiri (ks), Kuanua (ksd), Bafia (ksf), Kuot (kto), Kushi (kuh), Parkari Koli (kvx), Cornish (kw), Kwambi (kwm), Wadiyara Koli (kxp), Kyrgyz (ky), Rapoisi (kyx), Rangi (lag), Luxembourgish (lb), Tungag (lcm), Dũya (ldb), Ganda (lg), Ligurian (lij), Liberian English (lir), Kabras (lkb), Lala-Roba (lla), Lingala (ln), Longuda (lnu), Lao (lo), Loloda (loa), Loarki (lrk), Lasi (lss), Lithuanian (lt), Latgalian (ltg), Tsotso (lto), Luba-Lulua (lua), Luo (Kenya and Tanzania) (luo), Lushai (lus), Latvian (lv), Wanga (lwg), Yutanduchi Mixtec (mab), Mafa (maf), Maithili (mai), Huautla Mazatec (mau), North Moluccan Malay (max), Mbo (Cameroon) (mbo), Matsés (mcf), Masana (mcn), Mpiemo (mcx), Mbum (mdd), Maba (Chad) (mde), Moksha (mdf), Mekeo (mek), Meru (mer), Motu (meu), Marghi South (mfm), Cross River Mbembe (mfn), Mbe (mfo), Mandjak (mfv), Mpumpong (mgg), Lijili (mgi), Mungaka (mhk), Eastern Mari (mhr), Māori (mi), San Miguel El Grande Mixtec (mig), Cacaloxtepec Mixtec (miu), Macedonian (mk), Miya (mkf), Dhatki (mki), Malayalam (ml), Western Maninkakan (mlq), Mongolian (mn), Naba (mne), Manipuri (mni), Manggarai (mqy), Marathi (mr), Western Mari (mrj), Maria (India) (mrr), Marghi Central (mrt), Malay (ms), Musey (mse), Masikoro Malagasy (msh), Mansoanka (msw), Maltese (mt), Mewari (mtr), Tututepec Mixtec (mtu), Tidaá Mixtec (mtx), Mundang (mua), Musgu (mug), Musi (mui), Marwari (Pakistan) (mve), Indus Kohistani (mvy), Huitepec Mixtec (mxs), Mada (Cameroon) (mxu), Southeastern Nochixtlán Mixtec (mxy), Burmese (my), Erzya (myv), Mazatlán Mixe (mzl), Nalik (nal), Min Nan Chinese (nan), Neapolitan (nap), Norwegian Bokmål (nb), Ngamo (nbh), Notsi (ncf), Sibe (nco), Central Puebla Nahuatl (ncx), Samba Leko (ndi), Ndonga (ng), Ngizim (ngi), Tetelcingo Nahuatl (nhg), Zacatlán-Ahuacatlán-Tepetzintla Nahuatl (nhi), Central Nahuatl (nhn), Huaxcaleca Nahuatl (nhq), Nzanyi (nja), Dutch (nl), Ngombale (nla), Orizaba Nahuatl (nlv), Kwasio (nmg), Nawdm (nmz), Norwegian Nynorsk (nn), Ngiemboon (nnh), Norwegian (no), Nimadi (noe), Nepali (individual language) (npi), Pedi (nso), Nyanja (ny), Nyungwe (nyu), Occitan (oc), Od (odk), Odual (odu), Khana (ogo), Oromo (om), Orma (orc), Ormuri (oru), Odia (ory), Ossetic (os), Punjabi (pa), Central Pame (pbs), Southern Pashto (pbt), Northern Pashto (pbu), Nigerian Pidgin (pcm), Petats (pex), Phalura (phl), Pahari-Potwari (phr), Pero (pip), Piya-Kwonci (piy), Pökoot (pko), Polish (pl), Kohistani Shina (plk), Plateau Malagasy (plt), Northern Pame (pmq), Piemontese (pms), Papuan Malay (pmy), Western Panjabi (pnb), Poqomam (poc), San Juan Atzingo Popoloca (poe), San Felipe Otlaltepec Popoloca (pow), Ashéninka Perené (prq), Pashto (ps), Central Pashto (pst), Portuguese (pt), Western Highland Purepecha (pua), Paiwan (pwn), Chimborazo Highland Quichua (qug), Sipacapense (qum), Southern Pastaza Quechua (qup), Yanahuanca Pasco Quechua (qur), Santiago del Estero Quichua (qus), Sacapulteco (quv), Yauyos Quechua (qux), Ayacucho Quechua (quy), Ambo-Pasco Quechua (qva), Imbabura Highland Quichua (qvi), Loja Highland Quichua (qvj), Cajatambo North Lima Quechua (qvl), Corongo Ancash Quechua (qwa), Sihuas Ancash Quechua (qws), Chiquián Ancash Quechua (qxa), Puno Quechua (qxp), Santa Ana de Tusi Pasco Quechua (qxt), Arequipa-La Unión Quechua (qxu), Jauja Wanca Quechua (qxw), Logooli (rag), Romansh (rm), Romanian (ro), Tae' (rob), Rombo (rof), Rotokas (roo), Ratahan (rth), Russian (ru), Macedo-Romanian (rup), Kinyarwanda (rw), Sanskrit (sa), Yakut (sah), Santali (sat), Saleman (sau), Saya (say), Sindhi Bhil (sbn), Sardinian (sc), Shina (scl), Sicilian (scn), Sindhi (sd), Seri (sei), Chadian Arabic (shu), Sinhala (si), Sikkimese (sip), Siwai (siw), Siar-Lak (sjr), Slovak (sk), Sakalava Malagasy (skg), Saraiki (skr), Slovenian (sl), Shona (sn), Sinaugoro (snc), Soninke (snk), Somali (so), Solos (sol), Saposa (sps), Albanian (sq), Serbian (sr), Logudorese Sardinian (src), Campidanese Sardinian (sro), Sansi (ssi), Liana-Seti (ste), Sulka (sua), Swedish (sv), Svan (sva), Swahili (sw), Sakizaya (szy), Tamil (ta), Tangale (tan), Central Tarahumara (tar), Atayal (tay), Mandara (tbf), Malinaltepec Me'phaa (tcf), Tulu (tcy), Tondano (tdn), Tandroy-Mahafaly Malagasy (tdx), Telugu (te), Tajik (tg), Tigak (tgc), Thai (th), Chitwania Tharu (the), Kochila Tharu (thq), Rana Tharu (thr), Tahaggart Tamahaq (thv), Tigrinya (ti), Tigre (tig), Teop (tio), Turkmen (tk), Tesaka Malagasy (tkg), Kathoriya Tharu (tkt), Tlingit (tli), Filomena Mata-Coahuitlán Totonac (tlp), Tswana (tn), Toki Pona (tok), Tlacoapa Me'phaa (tpl), Tinputz (tpz), Tomoip (tqp), Turkish (tr), Kok Borok (trp), San Martín Itunyoso Triqui (trq), Sediq (trv), Torwali (trw), Tatar (tt), Tooro (ttj), Tera (ttr), Torau (ttu), Tupuri (tui), Tula (tul), Tedaga (tuq), Turkana (tuv), Tugen (tuy), Tidore (tvo), Tunen (tvu), Akan (tw), Termanu (twu), Tonsea (txs), Tanosy Malagasy (txy), Wuzlam (udl), Uyghur (ug), Ukrainian (uk), Kui (India) (uki), Umbundu (umb), Urdu (ur), Ushojo (ush), Uzbek (uz), Northern Uzbek (uzn), Vai (vai), Huarijio (var), Mom Jango (ver), Vietnamese (vi), Juxtlahuaca Mixtec (vmc), Ixtayutla Mixtec (vmj), Mitlatongo Mixtec (vmm), Soyaltepec Mazatec (vmp), Mazatlán Mazatec (vmz), Votic (vot), Võro (vro), Wakhi (wbl), Waci Gbe (wci), Wemale (weo), Cameroon Pidgin (wes), Waja (wja), Warji (wji), Wolof (wo), Gambian Wolof (wof), Xhosa (xh), Khetrani (xhe), Kalkoti (xka), Mingrelian (xmf), Antankarana Malagasy (xmv), Tsimihety Malagasy (xmw), Liberia Kpelle (xpe), Sinicahua Mixtec (xti), Cuyamecalco Mixtec (xtu), Yaqui (yaq), Yangben (yav), Agwagwune (yay), Eastern Yiddish (ydd), Yidgha (ydg), Tarok (yer), Nyankpa (yes), Yiddish (yi), Yoruba (yo), Yue Chinese (yue), Kinga (zga), Standard Moroccan Tamazight (zgh), Chinese (zh), Copainalá Zoque (zoc), Chimalapa Zoque (zoh), Rayón Zoque (zor), Chichicapan Zapotec (zpv), Mazaltepec Zapotec (zpy), Xanaguía Zapotec (ztg), Santa Catarina Albarradas Zapotec (ztn), Loxicha Zapotec (ztp), Tilquiapan Zapotec (zts), Güilá Zapotec (ztu), Zulu (zu), Zaza (zza)	Delivery instructions Voice design Emotion in descriptions or directions Speed · Duration · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `seednum_inference_stepsguidance_scalespeedduration_sect_shiftdenoisepreprocess_promptpostprocess_outputlayer_penalty_factorposition_temperatureclass_temperatureaudio_chunk_duration_secaudio_chunk_threshold_sectext_chunk_sizetext_chunk_mode` Vocal events laugh, sigh	CC-BY-NC-4.0 Model card and terms audio.cpp
OuteTTSaudio.cppOptional reference Exact variants (1) `outetts_1_0_1b_q8_0`q8_0	Documented languages Arabic, Belarusian, Bangla, German, English +18 Spanish (es), Persian (fa), French (fr), Hungarian (hu), Italian (it), Japanese (ja), Georgian (ka), Korean (ko), Lithuanian (lt), Latvian (lv), Dutch (nl), Polish (pl), Portuguese (pt), Russian (ru), Swahili (sw), Tamil (ta), Ukrainian (uk), Chinese (zh)	Sampling · Text chunking · Seed Generation parameters `max_tokenstemperaturetop_ktop_pmin_prepetition_penaltyrepetition_windowseedtext_chunk_sizetext_chunk_mode`	CC-BY-NC-SA-4.0 Model card and terms audio.cpp
Piper · Lessacaudio.cppPreset voice Exact variants (1) `piper_lessac_medium_orig`orig	Documented languages English	Speed · Text chunking · Seed Generation parameters `speedvariationduration_variationseedtext_chunk_modetext_chunk_size`	Voice / dataset-specific terms Model card and terms audio.cpp
Pocket TTSaudio.cppHybrid mode Exact variants (13) `pocket_tts_english_bf16`bf16 English Hybrid mode `pocket_tts_english_q8_0`q8_0 English Hybrid mode `pocket_tts_english_safetensors`native English Hybrid mode `pocket_tts_french_bf16`bf16 French Hybrid mode `pocket_tts_french_q8_0`q8_0 French Hybrid mode `pocket_tts_german_bf16`bf16 German Hybrid mode `pocket_tts_german_q8_0`q8_0 German Hybrid mode `pocket_tts_italian_bf16`bf16 Italian Hybrid mode `pocket_tts_italian_q8_0`q8_0 Italian Hybrid mode `pocket_tts_portuguese_bf16`bf16 Portuguese Hybrid mode `pocket_tts_portuguese_q8_0`q8_0 Portuguese Hybrid mode `pocket_tts_spanish_bf16`bf16 Spanish Hybrid mode `pocket_tts_spanish_q8_0`q8_0 Spanish Hybrid mode	By variant German, English, Spanish, French, Italian +1 Portuguese (pt)	Sampling · Text chunking · Seed Generation parameters `frames_after_eosmax_stepsmax_tokenstemperatureseednoise_clampeos_thresholdtruncate_clone_audiotext_chunk_size`	CC-BY-4.0 Model card and terms audio.cpp
Qwen3 TTSaudio.cppReference voice · Preset voice · Voice design Exact variants (10) `qwen3_tts_0_6b_base_bf16`bf16 Reference voice `qwen3_tts_0_6b_base_q8_0`q8_0 Reference voice `qwen3_tts_1_7b_base_bf16`bf16 Reference voice Variant source `qwen3_tts_1_7b_base_orig`orig Reference voice Variant source `qwen3_tts_1_7b_base_q8_0`q8_0 Reference voice Variant source `qwen3_tts_1_7b_base_safetensors`native Reference voice Variant source `qwen3_tts_1_7b_customvoice_bf16`bf16 Preset voice Variant source `qwen3_tts_1_7b_customvoice_q8_0`q8_0 Preset voice Variant source `qwen3_tts_1_7b_voicedesign_bf16`bf16 Voice design · Voice design Variant source `qwen3_tts_1_7b_voicedesign_q8_0`q8_0 Voice design · Voice design Variant source	Documented languages German, English, Spanish, French, Italian +5 Japanese (ja), Korean (ko), Portuguese (pt), Russian (ru), Chinese (zh)	Controls vary by variant Compare variants Qwen3 TTS 12Hz 0.6B Base / Qwen3 TTS 12Hz 1.7B Base / Qwen3 TTS 12Hz 1.7B Base Original-Dtype GGUF / Qwen3 TTS 12Hz 1.7B Base Safetensors qwen3_tts_0_6b_base_bf16, qwen3_tts_0_6b_base_q8_0, qwen3_tts_1_7b_base_bf16, qwen3_tts_1_7b_base_orig, qwen3_tts_1_7b_base_q8_0, qwen3_tts_1_7b_base_safetensors Sampling · Text chunking · Seed Generation parameters `seedmax_tokensdo_sampletemperaturetop_ktop_prepetition_penaltysubtalker_do_samplesubtalker_temperaturesubtalker_top_ksubtalker_top_ptext_chunk_size` Qwen3 TTS 12Hz 1.7B CustomVoice qwen3_tts_1_7b_customvoice_bf16, qwen3_tts_1_7b_customvoice_q8_0 Delivery instructions Emotion in descriptions or directions Sampling · Text chunking · Seed Generation parameters `seedmax_tokensdo_sampletemperaturetop_ktop_prepetition_penaltysubtalker_do_samplesubtalker_temperaturesubtalker_top_ksubtalker_top_ptext_chunk_size` Qwen3 TTS 12Hz 1.7B VoiceDesign qwen3_tts_1_7b_voicedesign_bf16, qwen3_tts_1_7b_voicedesign_q8_0 Voice description Voice design Emotion in descriptions or directions Sampling · Text chunking · Seed Generation parameters `seedmax_tokensdo_sampletemperaturetop_ktop_prepetition_penaltysubtalker_do_samplesubtalker_temperaturesubtalker_top_ksubtalker_top_ptext_chunk_size`	Apache-2.0 Model card and terms audio.cpp
sanoTTSaudio.cppPreset voice Exact variants (18) `sanotts_amy_orig`orig English Preset voice `sanotts_cs_orig`orig Czech Preset voice `sanotts_de_orig`orig German Preset voice `sanotts_es_orig`orig Spanish Preset voice `sanotts_fr_orig`orig French Preset voice `sanotts_heart_nano_orig`orig English Preset voice `sanotts_heart_orig`orig English Preset voice `sanotts_hfc_orig`orig English Preset voice `sanotts_hi_orig`orig Hindi Preset voice `sanotts_id_orig`orig Indonesian Preset voice `sanotts_it_orig`orig Italian Preset voice `sanotts_kristin_orig`orig English Preset voice `sanotts_ne_orig`orig Nepali Preset voice `sanotts_pt_orig`orig Portuguese Preset voice `sanotts_ro_orig`orig Romanian Preset voice `sanotts_ru_orig`orig Russian Preset voice `sanotts_tr_orig`orig Turkish Preset voice `sanotts_vi_orig`orig Vietnamese Preset voice	By variant Czech, German, English, Spanish, French +9 Hindi (hi), Indonesian (id), Italian (it), Nepali (ne), Portuguese (pt), Romanian (ro), Russian (ru), Turkish (tr), Vietnamese (vi)	Controls vary by variant Compare variants sanoTTS amy 1.46M FP32 GGUF (English, piperlite) / sanoTTS cs 1.57M FP32 GGUF (Czech, piperlite) / sanoTTS de 1.57M FP32 GGUF (German, piperlite) / sanoTTS es 1.56M FP32 GGUF (Spanish, piperlite) / sanoTTS fr 1.57M FP32 GGUF (French, piperlite) / sanoTTS hfc 1.83M FP32 GGUF (English, piperlite) / sanoTTS hi 1.50M FP32 GGUF (Hindi, piperlite) / sanoTTS id 1.56M FP32 GGUF (Indonesian, piperlite) / sanoTTS it 1.57M FP32 GGUF (Italian, piperlite) / sanoTTS kristin 1.40M FP32 GGUF (English, piperlite) / sanoTTS ne 1.47M FP32 GGUF (Nepali, piperlite) / sanoTTS pt 1.57M FP32 GGUF (Portuguese (Brazil), piperlite) / sanoTTS ro 1.57M FP32 GGUF (Romanian, piperlite) / sanoTTS ru 1.57M FP32 GGUF (Russian, piperlite) / sanoTTS tr 1.56M FP32 GGUF (Turkish, piperlite) / sanoTTS vi 1.57M FP32 GGUF (Vietnamese, piperlite) sanotts_amy_orig, sanotts_cs_orig, sanotts_de_orig, sanotts_es_orig, sanotts_fr_orig, sanotts_hfc_orig, sanotts_hi_orig, sanotts_id_orig, sanotts_it_orig, sanotts_kristin_orig, sanotts_ne_orig, sanotts_pt_orig, sanotts_ro_orig, sanotts_ru_orig, sanotts_tr_orig, sanotts_vi_orig Speed · Text chunking Generation parameters `speaking_ratetext_chunk_sizetext_chunk_mode` sanoTTS heart-nano 294k FP32 GGUF / sanoTTS heart 2.27M FP32 GGUF sanotts_heart_nano_orig, sanotts_heart_orig Speed · Text chunking · Seed Generation parameters `speaking_ratetext_chunk_sizetext_chunk_modeseed`	Weight terms not verified Model card and terms audio.cpp
Silero TTSsileroPreset voice Exact variants (9) `v5_cis_base_nostress` Azerbaijani, Bashkir, Belarusian, Chuvash, Armenian, Georgian, Kabardian, Khakas, Kazakh, Kyrgyz, Moksha, Erzya, Russian, Yakut, Tajik, Tatar, Udmurt, Ukrainian, Uzbek, Kalmyk Preset voice `v5_cis_ext` Chuvash, Kazakh, Tatar, Ukrainian, Uzbek, Kalmyk Preset voice `v5_5_ru` Russian Preset voice `v3_en` English Preset voice `v3_en_indic` English (India) Preset voice `v3_de` German Preset voice `v3_es` Spanish Preset voice `v3_fr` French Preset voice `v3_indic` Bangla, Gujarati, Hindi, Kannada, Malayalam, Manipuri, Rajasthani, Tamil, Telugu Preset voice	By variant Azerbaijani, Bashkir, Belarusian, Bangla, Chuvash +29 German (de), English (en), English (India) (en-in), Spanish (es), French (fr), Gujarati (gu), Hindi (hi), Armenian (hy), Georgian (ka), Kabardian (kbd), Khakas (kjh), Kazakh (kk), Kannada (kn), Kyrgyz (ky), Moksha (mdf), Malayalam (ml), Manipuri (mni), Erzya (myv), Rajasthani (raj), Russian (ru), Yakut (sah), Tamil (ta), Telugu (te), Tajik (tg), Tatar (tt), Udmurt (udm), Ukrainian (uk), Uzbek (uz), Kalmyk (xal)	Speed · Word stress Generation parameters `speedsample_ratestress`	Terms by variant MIT v5_cis_base_nostress Model terms CC-BY-NC-SA-4.0 v5_cis_ext, v5_5_ru, v3_en, v3_en_indic, v3_de, v3_es, v3_fr, v3_indic Model terms Model source
Sopranoaudio.cppPreset voice Exact variants (2) `soprano_1_1_80m_bf16`bf16 `soprano_1_1_80m_q8_0`q8_0	Documented languages English	Sampling · Seed Generation parameters `max_tokenstemperaturetop_prepetition_penaltyeos_biasseed`	Apache-2.0 Model card and terms audio.cpp
Sopro v2 Turboaudio.cppReference voice Exact variants (2) `sopro_v2_turbo_f16`f16 `sopro_v2_turbo_safetensors`orig	Documented languages German, English, French, Portuguese	Duration · Sampling · Synthesis steps · Text chunking · Seed Generation parameters `temperaturetop_ptop_knum_inference_stepsmax_secondsmin_secondstext_chunk_sizeseed`	Apache-2.0 Model card and terms audio.cpp
Supertonicaudio.cppPreset voice Exact variants (4) `supertonic_3_f16`f16 `supertonic_3_orig`orig `supertonic_3_q8_0`q8_0 `supertonic_3_safetensors`native	Documented languages Arabic, Bulgarian, Czech, Danish, German +26 Greek (el), English (en), Spanish (es), Estonian (et), Finnish (fi), French (fr), Hindi (hi), Croatian (hr), Hungarian (hu), Indonesian (id), Italian (it), Japanese (ja), Korean (ko), Lithuanian (lt), Latvian (lv), Dutch (nl), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Slovak (sk), Slovenian (sl), Swedish (sv), Turkish (tr), Ukrainian (uk), Vietnamese (vi)	Speed · Synthesis steps · Text chunking · Seed Generation parameters `num_inference_stepsspeedseedtext_chunk_sizetext_chunk_mode`	OpenRAIL-M Model card and terms audio.cpp
Tencent AuKaudio.cppOptional reference Exact variants (6) `auk_base_f16`f16 Optional reference `auk_base_f32`f32 Optional reference `auk_base_q8_0`q8_0 Optional reference `auk_flash_f16`f16 Optional reference Variant source `auk_flash_f32`f32 Optional reference Variant source `auk_flash_q8_0`q8_0 Optional reference Variant source	Language list not published	Controls vary by variant Compare variants AuK Base auk_base_f16, auk_base_f32, auk_base_q8_0 Voice description Duration · Guidance strength · Synthesis steps · Seed Generation parameters `duration_secnum_inference_stepsguidance_scalesway_sampling_coefseed` AuK-Flash auk_flash_f16, auk_flash_f32, auk_flash_q8_0 Voice description Duration · Seed Generation parameters `duration_secseed`	MIT Model card and terms audio.cpp
VibeVoiceaudio.cppPreset voice Exact variants (3) `vibevoice_1_5b_bf16`bf16 `vibevoice_1_5b_q8_0`q8_0 `vibevoice_7b_q8_0`q8_0 Variant source	Documented languages English, Chinese	Sampling · Guidance strength · Synthesis steps · Seed Generation parameters `num_inference_stepsguidance_scalemax_length_timesmax_tokensdo_sampletemperaturetop_ktop_pseed`	Terms by variant MIT vibevoice_1_5b_bf16, vibevoice_1_5b_q8_0 Model terms Converted weight terms not verified vibevoice_7b_q8_0 Model terms Model card and terms audio.cpp
VoiceTutaudio.cppOptional reference · Voice design Exact variants (2) `voicetut_tts_f16`f16 `voicetut_tts_q8_0`q8_0	Documented languages Egyptian Arabic	Delivery instructions Voice design Emotion in descriptions or directions Speed · Duration · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `seednum_inference_stepsguidance_scalespeedduration_sect_shiftdenoisepreprocess_promptpostprocess_outputlayer_penalty_factorposition_temperatureclass_temperatureaudio_chunk_duration_secaudio_chunk_threshold_sectext_chunk_sizetext_chunk_mode` Vocal events laugh, sigh	Apache-2.0 fine-tune; separate base and tokenizer terms Model card and terms audio.cpp
VoxCPMaudio.cppReference voice Exact variants (1) `voxcpm1_0_5b_q8_0`q8_0	Partial list English, Chinese	Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `text_chunk_modeseedmax_tokensmin_tokensnum_inference_stepsguidance_scaleretry_badcaseretry_badcase_max_timesretry_badcase_ratio_threshold`	Apache-2.0 Model card and terms audio.cpp
VoxCPM2audio.cppOptional reference · Voice design Exact variants (4) `voxcpm2_bf16`bf16 `voxcpm2_orig`orig `voxcpm2_q8_0`q8_0 `voxcpm2_safetensors`native	Partial list Arabic, Danish, German, Greek, English +25 Spanish (es), Finnish (fi), French (fr), Hebrew (he), Hindi (hi), Indonesian (id), Italian (it), Japanese (ja), Khmer (km), Korean (ko), Lao (lo), Malay (ms), Burmese (my), Dutch (nl), Norwegian (no), Polish (pl), Portuguese (pt), Russian (ru), Swedish (sv), Swahili (sw), Thai (th), Filipino (tl), Turkish (tr), Vietnamese (vi), Chinese (zh)	Delivery instructions Voice design Emotion in descriptions or directions Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `min_tokensmax_tokensnum_inference_stepsguidance_scaleretry_badcaseretry_badcase_max_timesretry_badcase_ratio_thresholdseedchunk_strategytext_chunk_sizetext_chunk_mode`	Apache-2.0 Model card and terms audio.cpp
Voxtral TTSvoxtralPreset voice Backend choices (3) These are backend selectors. The exact weights depend on the configured server. `auto` `gguf` `bf16`	Source claim Arabic, German, English, Spanish, French +4 Hindi (hi), Italian (it), Dutch (nl), Portuguese (pt)	Model defaults	CC-BY-NC-4.0 Model card and terms
Voxtral TTSaudio.cppPreset voice Exact variants (2) `voxtral_4b_tts_2603_bf16`bf16 `voxtral_4b_tts_2603_q8_0`q8_0	Documented languages Arabic, German, English, Spanish, French +4 Hindi (hi), Italian (it), Dutch (nl), Portuguese (pt)	Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `max_tokensnum_inference_stepsguidance_scaleseedtext_chunk_sizetext_chunk_mode`	CC-BY-NC-4.0 Model card and terms audio.cpp
XTTS v2xttsReference voice Exact variants (1) `tts_models/multilingual/multi-dataset/xtts_v2`	Source claim Arabic, Czech, German, English, Spanish +12 French (fr), Hindi (hi), Hungarian (hu), Italian (it), Japanese (ja), Korean (ko), Dutch (nl), Polish (pl), Portuguese (pt), Russian (ru), Turkish (tr), Simplified Chinese (zh-hans)	Speed Generation parameters `speed`	Coqui Public Model License 1.0.0 Model terms Model source
ZipVoice Distillaudio.cppReference voice Exact variants (2) `zipvoice_distill_gguf`orig `zipvoice_distill_q8_0`q8_0	Documented languages English, Chinese	Speed · Guidance strength · Synthesis steps · Text chunking · Seed Generation parameters `text_chunk_sizetext_chunk_modeguidance_scalenum_inference_stepst_shiftspeedfeat_scaletarget_rmsseed`	Apache-2.0 Model card and terms audio.cpp

Instructions and emotions are guidance. Check a short sample. Model-weight terms are separate from the Pandrator and audio.cpp software licences.

Transcription and word timing

Language recognition, word timing and speaker diarization are different capabilities. For transcription APIs, check the provider’s current catalogue.

Model	Listed languages	Timing and notes
Whisper large-v3	100 Show languages Afrikaans (af), Amharic (am), Arabic (ar), Assamese (as), Azerbaijani (az), Bashkir (ba), Belarusian (be), Bulgarian (bg), Bangla (bn), Tibetan (bo), Breton (br), Bosnian (bs), Catalan (ca), Czech (cs), Welsh (cy), Danish (da), German (de), Greek (el), English (en), Spanish (es), Estonian (et), Basque (eu), Persian (fa), Finnish (fi), Faroese (fo), French (fr), Galician (gl), Gujarati (gu), Hausa (ha), Hawaiian (haw), Hebrew (he), Hindi (hi), Croatian (hr), Haitian Creole (ht), Hungarian (hu), Armenian (hy), Indonesian (id), Icelandic (is), Italian (it), Japanese (ja), Javanese (jw), Georgian (ka), Kazakh (kk), Khmer (km), Kannada (kn), Korean (ko), Latin (la), Luxembourgish (lb), Lingala (ln), Lao (lo), Lithuanian (lt), Latvian (lv), Malagasy (mg), Māori (mi), Macedonian (mk), Malayalam (ml), Mongolian (mn), Marathi (mr), Malay (ms), Maltese (mt), Burmese (my), Nepali (ne), Dutch (nl), Norwegian Nynorsk (nn), Norwegian (no), Occitan (oc), Punjabi (pa), Polish (pl), Pashto (ps), Portuguese (pt), Romanian (ro), Russian (ru), Sanskrit (sa), Sindhi (sd), Sinhala (si), Slovak (sk), Slovenian (sl), Shona (sn), Somali (so), Albanian (sq), Serbian (sr), Sundanese (su), Swedish (sv), Swahili (sw), Tamil (ta), Telugu (te), Tajik (tg), Thai (th), Turkmen (tk), Filipino (tl), Turkish (tr), Tatar (tt), Ukrainian (uk), Urdu (ur), Uzbek (uz), Vietnamese (vi), Yiddish (yi), Yoruba (yo), Yue Chinese (yue), Chinese (zh)	Word timing through CrispASR; broad language coverage.
Parakeet TDT 0.6B v3	25 Show languages Bulgarian (bg), Croatian (hr), Czech (cs), Danish (da), Dutch (nl), English (en), Estonian (et), Finnish (fi), French (fr), German (de), Greek (el), Hungarian (hu), Italian (it), Latvian (lv), Lithuanian (lt), Maltese (mt), Polish (pl), Portuguese (pt), Romanian (ro), Slovak (sk), Slovenian (sl), Spanish (es), Swedish (sv), Russian (ru), Ukrainian (uk)	Native word timestamps for its listed languages, including Polish.
Qwen3 ASR 0.6B / 1.7B	30 Show languages Chinese (zh), English (en), Yue Chinese (yue), Arabic (ar), German (de), French (fr), Spanish (es), Portuguese (pt), Indonesian (id), Italian (it), Korean (ko), Russian (ru), Thai (th), Vietnamese (vi), Japanese (ja), Turkish (tr), Hindi (hi), Malay (ms), Dutch (nl), Swedish (sv), Danish (da), Finnish (fi), Polish (pl), Czech (cs), Filipino (fil), Persian (fa), Greek (el), Hungarian (hu), Macedonian (mk), Romanian (ro)	Timing is available for 20 languages through the aligner or Canary fallback. 10 recognition languages are transcript-only. Timed languages Chinese (zh), English (en), Yue Chinese (yue), German (de), French (fr), Spanish (es), Portuguese (pt), Italian (it), Korean (ko), Russian (ru), Japanese (ja), Dutch (nl), Swedish (sv), Danish (da), Finnish (fi), Polish (pl), Czech (cs), Greek (el), Hungarian (hu), Romanian (ro)
Qwen3 Forced Aligner 0.6B	11 Show languages Chinese (zh), English (en), Yue Chinese (yue), French (fr), German (de), Italian (it), Japanese (ja), Korean (ko), Portuguese (pt), Russian (ru), Spanish (es)	Alignment, rather than transcription. Its coverage is narrower than Qwen3-ASR.
MOSS Transcribe-Diarize 0.9B	50+ Model-card claim Languages named for the challenge English, French, German, Italian, Portuguese, Spanish, Japanese, Korean, Russian, Thai, Vietnamese, Filipino, Urdu, Turkish These are the 14 MLC-SLM challenge languages, rather than a complete support list.	Speaker-attributed timestamps with automatic language handling. Word alignment has narrower language coverage. The model owner does not publish a complete language list.

Qwen3-ASR recognises Polish; its native forced aligner does not support Polish. CrispASR uses a supported timing fallback. Not every ASR language has that option.

Input and output formats

Book/text input: TXT, PDF, EPUB, DOCX, MOBI, pasted text
Subtitle input: SRT; VTT through supported caption import paths
Audio/video input: WAV, MP3, FLAC, M4A, MP4, MKV, MOV, WebM and other supported media formats
Audio output: M4B, MP3, Opus, FLAC, WAV
Subtitle/text output: SRT, VTT, plain text; Quick Transcribe also offers JSON
Video output: MP4 with selectable or burned subtitles; narrated video-book output

DOCX and MOBI need Calibre. Scanned PDFs may need OCR. Local services have different memory and hardware requirements; check the selected package in the Manager.

Other local speech services

Alongside audio.cpp, the app has these service configurations. Some values are compatibility selectors rather than model-weight names.

Silero: v5_cis_base_nostress, v5_cis_ext, v5_5_ru, v3_en, v3_en_indic, v3_de, v3_es, v3_fr, v3_indic
Voxtral: auto, gguf, bf16
XTTS: tts_models/multilingual/multi-dataset/xtts_v2