Multilingual Speech Recognition on FLEURS (Language-Specific Results) benchmark leaderboard · SOTA2 Research

Multilingual Speech Recognition on FLEURS (Language-Specific Results)

2.47DE WERQwen3-Omni-30B-A3B-Instruct

Evaluation Results

Method	Links
Qwen3-Omni-30B-A3B-Instruct2026.07		2.47	5.11	—	—	2.34	—	3.07	19.85	—	—	3.35	—	1.55	—	—	3.15	—	—	—
Voxtral-Small-24B-25072026.07		2.58	5.45	—	—	2.77	—	3.71	17.1	—	—	3.54	—	2.16	—	—	6.29	—	—	—
Whisper-large-v32026.07		3.28	3.89	—	—	2.33	—	3.5	7.09	—	—	4.99	—	2.41	—	—	3.64	—	—	—
Qwen3-Omni-30B-A3B-Thinking2026.07		3.4	4.27	—	—	2.71	—	3.73	7.29	—	—	4.46	—	2.52	—	—	5.81	—	—	—
Canary-1B-v22026.07		3.47	—	—	—	2.63	—	4.19	—	—	—	4.41	—	2.35	—	—	5.92	—	—	—
Parakeet-TDT-0.6B-v32026.07		4.27	—	—	—	3.28	—	4.54	—	—	—	4.7	—	2.44	—	—	4.93	—	—	—
Audex 30B-A3B2026.07		4.33	5.11	—	—	2.85	—	4.35	9.71	—	—	5.79	—	3.27	—	—	5.45	—	—	—
Seamless-m4t-v2-large2026.07		5.29	6.77	—	—	3.99	—	6.94	14.76	—	—	6.37	—	3.52	—	—	6.51	—	—	—
Step-Audio-R1.12026.07		5.49	6.59	—	—	3.74	—	5.36	11.75	—	—	6.72	—	3.25	—	—	9.84	—	—	—
Audex 2B2026.07		5.98	6.38	—	—	3.71	—	4.93	10.22	—	—	6.89	—	4.22	—	—	8.74	—	—	—
AVLM-2B2026.06		8.44	13.45	6.5	18.71	6.18	45.5	5.47	15.97	11.79	10.15	9.44	11.69	4.66	16.98	18.3	8.09	20.35	11.25	12.7
Gemma-4-E2B-it2026.06		13.44	16.03	9.14	24.7	9.53	22.05	10.95	22.96	13.04	9.63	14.88	17.8	9.82	17.49	21.87	14.53	16.52	21.5	18.76
Qwen2.5-Omni-3B2026.06		28.9	41.15	4.18	50.33	33.71	70.3	36.38	56.07	27.46	54.97	22.58	37.94	14.79	36.02	67.14	64.17	75.96	5.33	54.54