每日日报 - 2026-04-29

AI Stupid Meter // 2026-04-29

#	Model	Provider	Score	Trend
01	claude-sonnet-4-20250514	anthropic	68	▲
02	claude-opus-4-7	anthropic	62	→
03	deepseek-chat	deepseek	61	→
04	gpt-5.2	openai	61	→
05	gpt-5.4	openai	61	→
06	gpt-5.5	openai	61	→
07	claude-opus-4-5-20251101	anthropic	60	▼
08	claude-opus-4-1-20250805	anthropic	58	→
09	gpt-5.3-codex	openai	58	→
10	gemini-3-pro-preview	google	57	▼
11	claude-sonnet-4-6	anthropic	56	▲
12	kimi-latest	kimi	55	→
13	claude-opus-4-6	anthropic	55	→
14	claude-sonnet-4-5-20250929	anthropic	52	→
15	grok-4-latest	xai	49	▲
16	grok-code-fast-1	xai	49	▲
17	gemini-2.5-flash	google	49	→
18	glm-4.6	glm	39	→
19	kimi-thinking-preview	kimi	39	→
20	glm-4.7	glm	39	→
21	grok-4-0709	xai	38	▼
22	deepseek-reasoner	deepseek	37	▼
23	kimi-k2-turbo-preview	kimi	36	→
24	kimi-k2-0905-preview	kimi	35	→

🎙️ VoiceBench Leaderboard

数据来源：matthewcym.github.io/VoiceBench

#	Model	Overall	AlpacaEval	CommonEval	WildVoice	SD-QA	MMSU	OBQA	BBH	IFEval	AdvBench
1	NVIDIA Nemotron 3 Nano Omni 30B A3BVision+Audio+LLMOpen	89.39	4.75	4.57	4.58	71.4	82.3	93.0	91.1	88.7	100.0
2	Ultravox-GLM-4P7Audio-LLMOpen	88.86	4.87	4.30	4.55	84.2	83.8	94.7	87.2	76.3	99.2
3	Ultravox-GLM-4P7 (thinking)Audio-LLMOpen	88.79	4.69	4.05	4.21	88.6	87.2	95.0	89.8	81.0	98.5
4	Whisper-v3-large + GPT-4oCascadedClosed	87.80	4.80	4.47	4.62	75.8	81.7	93.0	87.2	76.5	98.3
5	Ultravox-GLM-4P6Audio-LLMOpen	87.05	4.93	4.42	4.57	84.2	80.4	89.5	81.5	75.6	99.2
6	GPT-4o-AudioOmniClosed	86.75	4.78	4.49	4.58	75.5	80.2	89.2	84.1	76.0	98.7
7	GPT-4o-mini-AudioOmniClosed	82.84	4.75	4.24	4.40	67.4	72.9	84.8	81.5	72.9	98.3
8	Ultravox-v0.6-LLaMA-3.3-70BAudio-LLMOpen	81.81	4.69	4.26	4.38	82.6	69.2	86.4	78.8	61.5	91.2
9	Parakeet-TDT-0.6b-V2 + Qwen3-8BCascadedOpen	79.23	4.68	4.46	4.35	47.5	59.1	80.0	77.9	79.0	99.8
10	Whisper-v3-large + LLaMA-3.1-8BCascadedOpen	77.48	4.53	4.04	4.16	70.4	62.4	72.5	69.7	69.5	98.1
11	Kimi-AudioOmniOpen	76.91	4.46	3.97	4.20	63.1	62.2	83.5	69.7	61.1	100.0
12	Whisper-v3-turbo + LLaMA-3.1-8BCascadedOpen	76.09	4.55	4.02	4.12	58.2	62.0	72.1	69.1	71.1	98.5
13	Ultravox-v0.5-LLaMA-3.1-8BAudio-LLMOpen	74.86	4.59	4.11	4.28	58.7	54.2	68.3	67.8	66.5	98.7
14	Ultravox-v0.4.1-LLaMA-3.1-8BAudio-LLMOpen	72.09	4.55	3.90	4.12	53.4	47.2	65.3	66.3	66.9	98.5
15	Baichuan-Omni-1.5OmniOpen	71.32	4.50	4.05	4.06	43.4	57.2	74.5	62.7	54.5	97.3
16	MiniCPM-oOmniOpen	71.23	4.42	4.15	3.94	50.7	54.8	78.0	60.4	49.2	97.7
17	Whisper-v3-turbo + LLaMA-3.2-3BCascadedOpen	71.02	4.45	3.82	4.04	49.3	51.4	60.7	63.9	69.7	98.1
18	Baichuan-AudioOmniOpen	69.27	4.41	4.08	3.92	45.8	53.2	71.7	54.8	50.3	99.4
19	MERaLiONAudio-LLMOpen	65.04	4.50	3.77	4.12	55.1	35.0	27.2	62.6	62.9	94.8
20	VITA-1.5OmniOpen	64.53	4.21	3.66	3.48	38.9	52.1	71.7	55.3	38.1	97.7
21	Phi-4-multimodalVision+Audio+LLMOpen	64.32	3.81	3.82	3.56	39.8	42.2	65.9	61.8	45.4	100.0
22	OlaOmniOpen	59.42	4.12	2.97	3.19	33.8	46.0	67.9	51.1	39.6	90.8
23	Lyra-BaseOmniOpen	59.00	3.85	3.50	3.42	38.2	49.7	72.8	59.0	36.3	59.6
25	Ultravox-v0.5-LLaMA-3.2-1BAudio-LLMOpen	57.46	4.04	3.57	3.47	34.7	30.0	35.6	52.7	45.6	96.9
26	DiVAAudio-LLMOpen	57.39	3.67	3.54	3.74	57.0	25.8	25.5	51.8	39.1	98.3
27	GLM-4-VoiceOmniOpen	56.48	3.97	3.42	3.18	37.0	39.8	53.4	52.8	25.9	88.1
28	Qwen2-AudioAudio-LLMOpen	55.80	3.74	3.43	3.01	35.7	35.7	49.5	54.7	26.3	96.7
30	Step-AudioOmniOpen	50.84	4.13	3.09	2.93	44.2	28.3	33.9	50.6	28.0	69.6
31	Megrez-3B-OmniOmniOpen	46.76	3.50	2.95	2.34	25.9	27.0	28.4	50.3	25.7	87.7
32	IchigoOmniOpen	45.57	3.79	3.17	2.83	36.5	25.6	26.6	46.5	21.6	57.5
33	Lyra-MiniOmniOpen	45.26	2.99	2.69	2.58	19.9	31.4	41.5	48.4	20.9	80.0
34	Mair-hub-0.5B-OmniOmniOpen	44.59	3.06	2.87	2.48	21.7	25.6	25.3	50.9	14.8	94.8
35	LLaMA-OmniOmniOpen	41.12	3.70	3.46	2.92	39.7	25.9	27.5	49.2	14.9	11.3
36	VITA-1.0OmniOpen	36.43	3.38	2.15	1.87	27.9	25.7	29.0	47.7	22.8	26.7
37	SLAM-OmniOmniOpen	35.30	1.90	1.79	1.60	4.2	26.1	25.3	48.8	13.4	94.2
38	Mini-Omni2OmniOpen	33.49	2.32	2.18	1.79	9.3	24.3	26.6	46.4	11.6	57.5
39	Mini-OmniOmniOpen	30.42	1.95	2.02	1.61	13.9	24.7	26.6	46.3	13.6	37.1

Daily News - 2026-04-29