模型排行
1160 个模型 · 5 个榜单 · 2 个数据来源
语音王座
开源之光
国产之光
开源前三甲
Fish Audio S2 ProFish Audio · 总第 11 名1,121
Step Audio EditX (Mar 2026)StepFun · 总第 16 名1,101
Magpie-Multilingual 357M (Feb 2026)NVIDIA · 总第 23 名1,071
这一期的变化
GPT-5.5 (xhigh)智力榜 · 第 1新上榜
GPT Image 2 (high)图像榜 · 第 1新上榜
Realtime TTS 1.5 Max语音榜 · 第 1新上榜
Claude Opus 4.7 (Adaptive Reasoning, Max Effort)智力榜 · 第 2新上榜
kimi-k2.6编程榜 · 第 4新上榜
| 排名 | 变化 | 模型 | 得分 | 置信区间 |
|---|---|---|---|---|
| 1 | 新 | Realtime TTS 1.5 Max闭源Inworld | 1,208 | ±19 |
| 2 | — | Gemini 3.1 Flash TTS闭源Google | 1,204 | ±18 |
| 3 | — | Eleven v3闭源ElevenLabs | 1,176 | ±12 |
| 4 | — | Inworld TTS 1 Max闭源Inworld | 1,165 | ±19 |
| 5 | — | Speech 2.8 HD闭源MiniMax | 1,164 | ±12 |
| 6 | 新 | Realtime TTS 1.5 Mini闭源Inworld | 1,160 | ±16 |
| 7 | ▲ 2 | Step TTS 2 (Mar 2026)闭源StepFun | 1,150 | ±20 |
| 8 | — | Speech 2.8 Turbo闭源MiniMax | 1,146 | ±12 |
| 9 | ▲ 1 | Speech 2.6 HD闭源MiniMax | 1,132 | ±12 |
| 10 | ▲ 1 | Speech 2.6 Turbo闭源MiniMax | 1,129 | ±11 |
| 11 | ▼ 5 | Fish Audio S2 Pro开源Fish Audio | 1,121 | ±22 |
| 12 | ▲ 2 | Azure HD 2.5闭源Microsoft | 1,119 | ±22 |
| 13 | — | Speech-02-HD闭源MiniMax | 1,118 | ±11 |
| 14 | ▼ 2 | Inworld TTS 1闭源Inworld | 1,118 | ±17 |
| 15 | — | Multilingual v2闭源ElevenLabs | 1,107 | ±8 |
| 16 | — | Step Audio EditX (Mar 2026)开源StepFun | 1,101 | ±22 |
| 17 | — | Speech-02-Turbo闭源MiniMax | 1,101 | ±12 |
| 18 | — | TTS-1闭源OpenAI | 1,101 | ±8 |
| 19 | ▲ 1 | Turbo v2.5闭源ElevenLabs | 1,097 | ±8 |
| 20 | ▼ 1 | TTS-1 HD闭源OpenAI | 1,096 | ±12 |
| 21 | — | Flash v2.5闭源ElevenLabs | 1,087 | ±9 |
| 22 | — | Gemini 2.5 Flash Lite TTS闭源Google | 1,077 | ±14 |
| 23 | ▲ 6 | Magpie-Multilingual 357M (Feb 2026)开源NVIDIA · 另有 1 个档位 | 1,071 | ±24 |
| 24 | ▼ 1 | Sonic 3闭源Cartesia | 1,068 | ±13 |
| 25 | ▼ 1 | OpenAudio S1闭源Fish Audio | 1,064 | ±11 |
| 26 | ▼ 1 | SIMBA 1.6闭源Speechify | 1,062 | ±21 |
| 27 | ▼ 1 | Studio闭源Google | 1,061 | ±10 |
| 28 | ▼ 1 | T2A-01-HD闭源MiniMax | 1,059 | ±10 |
| 29 | ▼ 1 | Kokoro 82M v1.0开源Kokoro | 1,056 | ±9 |
| 30 | ▲ 1 | Polly Generative闭源Amazon | 1,054 | ±11 |
| 31 | ▲ 1 | AsyncFlow V2, async闭源Async | 1,053 | ±12 |
| 32 | ▲ 2 | Maya1开源Maya Research | 1,050 | ±13 |
| 33 | ▼ 3 | Voxtral TTS开源Mistral | 1,049 | ±23 |
| 34 | ▼ 1 | Azure Neural闭源Microsoft | 1,048 | ±8 |
| 35 | ▲ 2 | Sonic English (Oct 2024)闭源Cartesia | 1,045 | ±12 |
| 36 | ▼ 1 | Octave 2闭源Hume AI | 1,043 | ±14 |
| 37 | ▲ 1 | Polly Long-Form闭源Amazon | 1,042 | ±20 |
| 38 | ▼ 2 | Chirp 3: HD闭源Google | 1,041 | ±13 |
| 39 | ▲ 1 | Chatterbox HD闭源Resemble AI | 1,035 | ±10 |
| 40 | ▼ 1 | Gemini 2.5 Flash TTS (Dec 2025)闭源Google | 1,032 | ±15 |
| 41 | — | Journey闭源Google | 1,028 | ±18 |
| 42 | — | SIMBA 1.0闭源Speechify | 1,026 | ±10 |
| 43 | ▲ 1 | Gemini 2.5 Pro (Dec 2025)闭源Google | 1,024 | ±16 |
| 44 | ▼ 1 | MiMo-V2-TTS闭源Xiaomi | 1,023 | ±22 |
| 45 | — | T2A-01-Turbo闭源MiniMax | 1,021 | ±10 |
| 46 | — | Lightning v3.1闭源Smallest.ai | 1,021 | ±23 |
| 47 | ▲ 2 | MAI-Voice-1闭源Microsoft | 1,021 | ±22 |
| 48 | ▼ 1 | Octave TTS闭源Hume AI | 1,018 | ±10 |
| 49 | ▼ 1 | Fish Speech 1.5开源Fish Audio | 1,013 | ±10 |
| 50 | — | Chatterbox开源Resemble AI | 1,006 | ±10 |
智力进步
涨得最多 · Anthropic +57.2目前最高 · Anthropic 64.6最高与最低相差 18.1 分时间范围 2025年11月 – 2026年11月
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi