Porto

Here's the AI app I built to help me and my wife have kids. You can find a link to it in the next post.

@MarioNawfal @elonmusk @XFreeze @xai @grok is it different than whisper large v3? How does it compare?

🇺🇸 xAI just dropped a real-time speech-to-text model built for voice apps: high limits, multi-language support, the works. Priced at just $0.10–$0.20/hour, crushing most competitors. Quiet release, loud warning to the rest." Source: @XFreeze, @xAI, @elonmusk

@livekit Does it work for Portuguese ? @grok

Pronunciation is one of the fastest ways to break trust in a voice agent, especially in healthcare, legal, and finance where terminology matters. Rime's Mist v3 introduces phonetic brackets that let you define the exact pronunciation for any word and reproduce it deterministically. We built a demo nurse agent that stumbled on words like "levothyroxine" and "gastroesophageal," then fixed every one with a few lines of config. It's also fast.. as low as 100ms TTFB. Try it on LiveKit Inference today.

@Jota_DBS @grok find the full interview

Para cuando te digan que la Sábana Santa es una: "falsificación medieval que ya ha sido parcialmente replicada por la ciencia moderna"

@BroMbaku @mjtechguy @NousResearch I think that’s too small for any useful model

@mjtechguy @NousResearch I have an old laptop with 4GB VRAM, Which small model should I run it with hermes?!

Hermes agent by @NousResearch is elite. Setup was a breeze, integrations were super easy and it just seems to work. Way easier than openclaw was. Great work. Also, GLM-5-Turbo is cooking.

@XFreeze @grok does it speak in Portuguese (brazil)? Can it call tools?

xAI's Grok Voice Agent API is insanely good.. You get a voice agent that sounds completely human and handles your always-active customer support across any domain, working for you 24/7 You can now build full voice agents - Medical Office, Restaurant Host, Real Estate, Hotel Concierge, Help Desk or create your own from scratch Even you can do it without any coding or needing a team to spend months setting things up Anyone can now deploy a real AI voice agent to handle calls around the clock for just $0.05/min That's only $3/hour for a worker that never sleeps, never calls in sick, and seamlessly speaks dozens of languages This is the stuff that used to cost companies millions. xAI just made it a template

@MilksandMatcha Ive been using codex to advance fertility research, following natural conception methods (not IVF). I’ve been helping dozens of obgyn in Brazil by conducting a research project

Giving away 5 Codex Pro plans Each person will get 3 months of free Codex Pro (highest tier). Winners will be selected from comments in 48 hours, comment below why you want it.

@floriandarroman @grok anyone can buy this like airalo?

This guy is our French Elon Musk (even cooler than Elon, I would say) He just released a new mobile service that covers all of these red countries (doesn't mean they're communists) with unlimited data for €19.90/mo. Now I can just use my French number in Indonesia 🫶

@colmflynnire @grok is it permanent on Vatican? When was it placed?

One of my favourite pieces in St Peter’s Square at the Vatican: a statue that first appears to be a beggar, transforming into an angel as you move around it - by Timothy Schmalz.

@forgebitz 7

last year we had some massive billboards next to the highway/airport if you guess how many signups we got from it you get a free hat

@DrNeilStone @grok ?

Most antibiotic course durations are completely made up

@grok @psk90_ai @art_zucker What providers offer cohere transcribe? And what’s the cost?

Cohere Transcribe doesn't publish a public per-language WER for Brazilian Portuguese yet, but it beats Whisper Large v3 overall: 5.42% avg WER across 14 langs (incl. PT) vs Whisper's 7.44%. On the CER leaderboard you referenced, Cohere hits 16.46% vs Whisper Large v3 Turbo at 19.62%. For multilingual/BR-PT production use, Cohere is the stronger pick based on their evals (outperforms or matches best open models on the other 13 langs). Check their blog for details!

Cohere Transcribe is #2 on the Open ASR Leaderboard. But it's #1 if you need multilingual. ━━━━━━━━━━━━━━━━━━━ Parakeet 0.6B takes the top spot at 14.74% CER — but it's English + Japanese only. Cohere Transcribe sits at #2 (16.46% CER) with 14 languages, 2B params, Apache 2.0. For any enterprise shipping outside English — this is the model. → English, French, German, Spanish, Portuguese, Italian, Dutch, Arabic, Chinese, Japanese, Korean, Greek, Polish, Vietnamese → 161x real-time inference → Handles noisy real-world audio, not just clean benchmarks → vLLM support out of the box In production, you rarely get to pick your language. The model that wins is the one that works across all of them. 🔗 Model: lnkd.in/ggfeZye5 🔗 Blog: cohere.com/blog/transcribe Building multilingual voice AI for enterprise? DM me at Zingaro Ai and LiteCompute AII. ♻️ Repost if useful. Follow Pasha S for daily open-source AI drops.

@abracetradicao @nayibbukele @grok o que é britcard?

>construiu mega prisões >aprovou prisão perpétua para hom1cidi0, estupr0 e terr0rrismo >os crimes violentos diminuíram 95% >El Salvador passou de país mais perigoso para país mais seguro da América Latina >mais de 90% do povo aprova seu governo >disse que a batalha era espiritual e que venceram porque rezaram >chamou o Britcard de marca da besta >está retirando a poluição visual das ruas Você simplesmente pode fazer as coisas acontecerem.

@grok @agcardososp @fiscaldofim @grok mostre o link que prova que não foi carro autonomo

Isso aconteceu na Ponte Lusong, em Zhuzhou, província de Hunan, na China, em 23/09/2024 (07:36 local). O carro era um Zeekr elétrico branco (EV chinês de luxo). Ele foi atingido por trás por um táxi, perdeu o controle, invadiu a contramão e colidiu com vários veículos e motos. 6 mortos e 7 feridos. Não era modo autônomo total, mas há relatos de aceleração descontrolada pós-impacto.

Loucura total: um carro autônomo apresentou falha, invadiu a pista contrária e saiu acelerando descontroladamente, colidindo com vários veículos à frente…

@OfficialLoganK i decided to try again and it accepted research project name

hi @OfficialLoganK trying gemini api for the first time and i have some feedbacks. 1. I can't simply create a key, I have to create a project first 2. project name must have min-4char (why?) 3. I chose name `research` and got an error as the 'request was suspicious'

@fefestudyin vc tinha q usar o app 'dr. assistente' baixa na app store pra vc ver... n precisa digitar nada durante a consulta q ele faz td pra vc

gente acabei de ter uma experiência individual na faculdade kkkkkkkk no meio da consulta do paciente, do nada entra um cara falando q vai trocar o teclado do computador pq tava quebrado (letra Z n funcionava). até ai "ok" apesar do desconforto de fazer isso logo na frente do

@fortelabs Welcome home. Start by going to the mass every Sunday. I think you would like Opus Dei and Saint Josemaria Escrivá teachings. You should take a look with open heart.

After 20 years as an atheist, and slowly warming to religion, I find it bizarre that it's actually Catholicism of all things that most appeals to me But the Latin American version: all about family, taking many days off from work or school, throwing gigantic festivals, dancing nude in the streets, and having a patron saint who's always watching out for you

@xai please integrate with @ElevenLabs

Grok's Text to Speech API is now available. Start building with natural voices and expressive controls to bring your apps to life. #text-to-speech" target="_blank" rel="nofollow noopener">x.ai/api/voice#text…

@coproduto Na minha empresa eu deixo um cartão de crédito com 50usd pra cada pessoa pra cada um escolher o serviço q quiser usar. Se precisar de mais a gnt aumenta. Quero que todos usem, quanto mais aumentar a produtividade individual, melhor

Devs: A empresa onde vocês trabalham tá colocando meta pra uso de IA? Se sim, deram alguma orientação e/ou tooling pra vocês?