rime_coda_clementine_vs_elevenlabs_flash25_full_cmos_noemail_v2

Full CMOS benchmark: Rime Coda (clementine) vs ElevenLabs Flash 2.5 (apply_text_normalization=on), all corpus pairs except email addresses, 7 raters/pair. Alphanumerics conditioned: spell() for Rime, plain-text spelling standard for ElevenLabs; leading '#' dropped from all conditioned inputs. Both sides 24kHz mono 16-bit WAV.

Overall

rime_coda_clementine (Model B)

-0.42

elevenlabs_flash25_vchg6ski (Model A)

By Tag

rime_coda_clementine (Model B)

elevenlabs_flash25_vchg6ski (Model A)

Answers

2. elevenlabs_flash25_vchg6ski (Model A) is much better

-1

-2. rime_coda_clementine (Model B) is much better

517 (15.13%)

525 (15.37%)

528 (15.46%)

693 (20.29%)

1153 (33.75%)

Deep analysis

2. elevenlabs_flash25_vchg6ski (Model A) is much better

-1

-2. rime_coda_clementine (Model B) is much better

		Tags	Statistics
00:00 --:-- elevenlabs_flash25_vchg6ski/support_dialogue-0540.wav 00:00 --:-- rime_coda_clementine/support_dialogue-0540.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_support_dialogue sub_travel utt_support_dialogue-0540 campaign_full_cmos_v2 cat_support_dialogue sub_travel utt_support_dialogue-0540 campaign_full_cmos_v2	0.57 1.18 0.48
00:00 --:-- elevenlabs_flash25_vchg6ski/alphanumerics-0130.wav 00:00 --:-- rime_coda_clementine/alphanumerics-0130.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_alphanumerics sub_flight_number utt_alphanumerics-0130 campaign_full_cmos_v2 cat_alphanumerics sub_flight_number utt_alphanumerics-0130 campaign_full_cmos_v2	-0.86 1.55 0.63
00:00 --:-- elevenlabs_flash25_vchg6ski/phone_numbers-0104.wav 00:00 --:-- rime_coda_clementine/phone_numbers-0104.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_phone_numbers sub_extension utt_phone_numbers-0104 campaign_full_cmos_v2 cat_phone_numbers sub_extension utt_phone_numbers-0104 campaign_full_cmos_v2	0.29 1.48 0.61
00:00 --:-- elevenlabs_flash25_vchg6ski/phone_numbers-0023.wav 00:00 --:-- rime_coda_clementine/phone_numbers-0023.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_phone_numbers sub_toll_free utt_phone_numbers-0023 campaign_full_cmos_v2 cat_phone_numbers sub_toll_free utt_phone_numbers-0023 campaign_full_cmos_v2	-1.86 0.35 0.14
00:00 --:-- elevenlabs_flash25_vchg6ski/dates_times-0051.wav 00:00 --:-- rime_coda_clementine/dates_times-0051.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_dates_times sub_date utt_dates_times-0051 campaign_full_cmos_v2 cat_dates_times sub_date utt_dates_times-0051 campaign_full_cmos_v2	1.00 1.07 0.44
00:00 --:-- elevenlabs_flash25_vchg6ski/support_dialogue-0133.wav 00:00 --:-- rime_coda_clementine/support_dialogue-0133.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_support_dialogue sub_healthcare utt_support_dialogue-0133 campaign_full_cmos_v2 cat_support_dialogue sub_healthcare utt_support_dialogue-0133 campaign_full_cmos_v2	0.14 1.64 0.67
00:00 --:-- elevenlabs_flash25_vchg6ski/alphanumerics-0106.wav 00:00 --:-- rime_coda_clementine/alphanumerics-0106.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_alphanumerics sub_policy_case_id utt_alphanumerics-0106 campaign_full_cmos_v2 cat_alphanumerics sub_policy_case_id utt_alphanumerics-0106 campaign_full_cmos_v2	-0.57 1.68 0.69
00:00 --:-- elevenlabs_flash25_vchg6ski/support_dialogue-0026.wav 00:00 --:-- rime_coda_clementine/support_dialogue-0026.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_support_dialogue sub_banking utt_support_dialogue-0026 campaign_full_cmos_v2 cat_support_dialogue sub_banking utt_support_dialogue-0026 campaign_full_cmos_v2	-0.71 1.28 0.52

Overall

rime_coda_clementine (Model B)

-0.42

elevenlabs_flash25_vchg6ski (Model A)

By Tag

rime_coda_clementine (Model B)

elevenlabs_flash25_vchg6ski (Model A)

Answers

2. elevenlabs_flash25_vchg6ski (Model A) is much better

-1

-2. rime_coda_clementine (Model B) is much better

517 (15.13%)

525 (15.37%)

528 (15.46%)

693 (20.29%)

1153 (33.75%)

Deep analysis

2. elevenlabs_flash25_vchg6ski (Model A) is much better

-1

-2. rime_coda_clementine (Model B) is much better

		Tags	Statistics
00:00 --:-- elevenlabs_flash25_vchg6ski/support_dialogue-0540.wav 00:00 --:-- rime_coda_clementine/support_dialogue-0540.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_support_dialogue sub_travel utt_support_dialogue-0540 campaign_full_cmos_v2 cat_support_dialogue sub_travel utt_support_dialogue-0540 campaign_full_cmos_v2	0.57 1.18 0.48
00:00 --:-- elevenlabs_flash25_vchg6ski/alphanumerics-0130.wav 00:00 --:-- rime_coda_clementine/alphanumerics-0130.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_alphanumerics sub_flight_number utt_alphanumerics-0130 campaign_full_cmos_v2 cat_alphanumerics sub_flight_number utt_alphanumerics-0130 campaign_full_cmos_v2	-0.86 1.55 0.63
00:00 --:-- elevenlabs_flash25_vchg6ski/phone_numbers-0104.wav 00:00 --:-- rime_coda_clementine/phone_numbers-0104.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_phone_numbers sub_extension utt_phone_numbers-0104 campaign_full_cmos_v2 cat_phone_numbers sub_extension utt_phone_numbers-0104 campaign_full_cmos_v2	0.29 1.48 0.61
00:00 --:-- elevenlabs_flash25_vchg6ski/phone_numbers-0023.wav 00:00 --:-- rime_coda_clementine/phone_numbers-0023.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_phone_numbers sub_toll_free utt_phone_numbers-0023 campaign_full_cmos_v2 cat_phone_numbers sub_toll_free utt_phone_numbers-0023 campaign_full_cmos_v2	-1.86 0.35 0.14
00:00 --:-- elevenlabs_flash25_vchg6ski/dates_times-0051.wav 00:00 --:-- rime_coda_clementine/dates_times-0051.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_dates_times sub_date utt_dates_times-0051 campaign_full_cmos_v2 cat_dates_times sub_date utt_dates_times-0051 campaign_full_cmos_v2	1.00 1.07 0.44
00:00 --:-- elevenlabs_flash25_vchg6ski/support_dialogue-0133.wav 00:00 --:-- rime_coda_clementine/support_dialogue-0133.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_support_dialogue sub_healthcare utt_support_dialogue-0133 campaign_full_cmos_v2 cat_support_dialogue sub_healthcare utt_support_dialogue-0133 campaign_full_cmos_v2	0.14 1.64 0.67
00:00 --:-- elevenlabs_flash25_vchg6ski/alphanumerics-0106.wav 00:00 --:-- rime_coda_clementine/alphanumerics-0106.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_alphanumerics sub_policy_case_id utt_alphanumerics-0106 campaign_full_cmos_v2 cat_alphanumerics sub_policy_case_id utt_alphanumerics-0106 campaign_full_cmos_v2	-0.57 1.68 0.69
00:00 --:-- elevenlabs_flash25_vchg6ski/support_dialogue-0026.wav 00:00 --:-- rime_coda_clementine/support_dialogue-0026.wav	elevenlabs_flash25_vchg6ski rime_coda_clementine	cat_support_dialogue sub_banking utt_support_dialogue-0026 campaign_full_cmos_v2 cat_support_dialogue sub_banking utt_support_dialogue-0026 campaign_full_cmos_v2	-0.71 1.28 0.52