Smallest AI's Pulse beats Sarvam and Deepgram on Hindi STT benchmark
vaibhavbetter · x · 2026-11-02
Smallest AI tested three speech-to-text models against Shankar Mahadevan's ultra-fast "Breathless"—two gave up, its Pulse handled it. On real Hindi contact-center calls, Pulse posted a 10.20% word error rate, lower than Sarvam Saaras v4 and Deepgram Nova 3.
More from Models
- Unverified: Google quietly rolling out 'Gemini 4 Carbon' to cyber defenders and partners — thione · 2026-11-02
- Unverified: Mistral opens 1T-param Large 4 weights; Anthropic faces $41M Elvis songs lawsuit — thione · 2026-11-02
- Mistral employee shows its model autonomously picked STO-3G basis set, hit 1e-13 accuracy vs PySCF — MistralAI · 2026-11-02
- Fable 5.5 launch week teased: "see you real soon" — iruletheworldmo · 2026-11-02
- Amazon CEO expects just 6-8 AI models to be 'good enough' for most enterprise work — YvesMulkers · 2026-11-02
- ZenteiQ, backed by IndiaAI Mission, to unveil homegrown foundation models tomorrow — deepakns · 2026-11-02