Comments (15)

• @theo_avramiss ♥ 36 2026-06-18

  So can someone use this on low level hardware? Like without a GPU?

• @jasper_de_craene ♥ 35 2026-06-19

  its really good. recognizes multiple languages, even when switching language mid sentence.

• @marc.kaz ♥ 19 2026-06-18

  🚨 NVIDIA JUST DROPPED A 0.6B CPU-ONLY SPEECH RECOGNITION MODEL
  
  Nemotron-3.5-ASR delivers real-time streaming ASR across 40+ languages — and it runs entirely on CPU.
  
  Ships with:
  • Only 0.6B parameters
  • Real-time streaming output
  • 2.5x faster than official Nemo runtime (same accuracy)
  • Fully offline capable
  • Easy integration into local agent pipelines
  
  Another strong small model for on-device/local AI stacks.
  
  👉 https://huggingface.co/nvidia/nemotron-3.5-asr-streaming-0.6b
  
  Who’s running local speech recognition today? Drop a 🔥

• @shahar_ain_kedem ♥ 15 2026-06-18

  You are the best Ai influencer

• @rohantg_ ♥ 9 2026-06-18

  how is the goat finding these 🔥🙌

• @jj_elly.fish ♥ 7 2026-06-18

  Brooo check Mothrag.com new opensource commercial multihop retrival. 🫪😱

• @sami.p_1707 ♥ 5 2026-06-19

  nemotron 3.5 vs whisper?

• @cquin__ ♥ 3 2026-06-18

  Foh-ee languages

• @benitserge ♥ 3 2026-06-18

  WhatsApp guyz🔥🔥🔥👏 you have hot news everytime

• @llamapalhq ♥ 2 2026-06-19

  Yo @marc.kaz ! This is actually a good model. I added it to my app LlamaPal. Do check it out. Its on Playstore and it's free.

• @arnabwithab ♥ 2 2026-06-19

  What is that 241 ms? If thats the actual latency, im not sure its worth it

• @suharshh ♥ 2 2026-06-19

  @sami.p_1707 its working better than whisper for other languages (tested hindi)

• @nixkuroi ♥ 2 2026-06-19

  How does it compare to Whisper large model?

• @louxhist ♥ 2 2026-06-18

  you guys have tried it, how good it's comparable to grow super whisper?

• @tyl0o0r ♥ 1 2026-06-18

  Why is this any better than existing non NN text to speech tools?