Krishna Gaurav

Building things robustly. Breaking tech news daily. Sharing what actually works so you don't have to learn the hard way.

[Krishna Gaurav @krigrv]

@TheTempleBoy Agreed

[Krishna Gaurav @krigrv]

@Migma_AI hey there :)

[Krishna Gaurav @krigrv]

@VittoStack would love to join it

[Krishna Gaurav @krigrv]

@elder_plinius what do we do?

[Krishna Gaurav @krigrv]

Claude just released Fable 5.

Mythos-class model. Previously held back on safety grounds. Now cleared for general use.

They say it's the most capable model they've ever shipped publicly. No caveats.

@claudeai

[Claude @claudeai]

Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.

Its capabilities exceed those of any model we’ve ever made generally available.

[Krishna Gaurav @krigrv]

NotebookLM just got agentic.

Chat now does multi-step research on its own. Better reasoning. New output formats. Not just summarizing your PDFs anymore.

Rolling out to Google AI Ultra subscribers.

@NotebookLM

[Gemini Notebook @Gemini_Notebook]

Introducing a more powerful NotebookLM 🚀

Massive upgrades deliver agentic capabilities in chat, more advanced reasoning, and a suite of new output formats. Tackling complex, multi-step research problems has never been easier.

Rolling out now to Google AI Ultra subscribers.

[Krishna Gaurav @krigrv]

Apodex just shipped 2.0.

A multi-agent deep research team. Searches the web, reasons over evidence, writes reports.

Every claim in the report links back to an evidence chain. Independently audited before you see the output.

SOTA on research benchmarks. Not vibe-based summaries.

@Apodex_AI

[Apodex @Apodex_AI]

Meet 𝗔𝗽𝗼𝗱𝗲𝘅 𝟭.𝟬 🔭 — a heavy-duty agent team for deep research, which sets the SOTA! The team searches the web, reasons over evidence, and writes reports where every claim is backed by an explicit 𝘦𝘷𝘪𝘥𝘦𝘯𝘤𝘦 𝘤𝘩𝘢𝘪𝘯, independently audited before delivery.
🌐 https://t.co/pOQAjL92uF

[Krishna Gaurav @krigrv]

Google just shipped Gemini 3.5 Live Translate.

Speech-to-speech. 70+ languages. Starts translating the moment you open your mouth.

Streams continuously — listens to what you're saying next while delivering what you already said. No pauses. No "wait for it."

Real-time bilingual conversation without that awkward silence between sentences.

@GoogleAI

[Google AI @GoogleAI]

Today, we released Gemini 3.5 Live Translate, our latest audio model for live speech-to-speech translation.

It supports over 70 languages and starts translating as soon as you start talking, streaming translations while listening to what you say next. No awkward pauses or choppy audio, just real connection without language barriers.

So, how does it work? 🤔

The model is able to make split-second decisions to juggle speed and translation quality so conversations actually feel fluid, human, and natural. In order to do this, the model must receive and contextualize the input while simultaneously outputting the translated speech.

Through this process, Gemini 3.5 Live Translate manages to stay mere seconds behind each speaker and can even maintain pacing, pitch, and intonation across extended sessions.

See it in action below, or try it yourself in the Google Translate app on iOS & Android.

[Krishna Gaurav @krigrv]

Kimi just dropped Kimi Work. A local desktop AI agent.

Up to 300 agents running in parallel on your machine. Everything stays local.

Paired with a browser extension (WebBridge) that lets agents navigate websites naturally — search, scroll, click.

No cloud. No API bills. Just your Mac/PC doing the work.

Not a demo. Ships with a native agent swarm.

@Kimi_Moonshot

[Kimi.ai @Kimi_Moonshot]

Meet Kimi Work - a local AI agent on your desktop that does the work for you.

🔹Native agent swarm: Up to 300 AI agents running in parallel on your local machine.
🔹Browser use: Paired with WebBridge extension, your agent will navigate websites in your browser: search, scroll, click, type and complete tasks.
🔹Built for Finance: Native global market data tool call from Yahoo Finance and World Bank - no complex API setup required.
🔹Memory system: Kimi Desktop keeps a running diary of your preferences, past decisions, and context to know you better.

Available for macOS (Apple Silicon) and Windows.

🔗Try it now: https://t.co/yhiai2VWIy

[Google AI Developers @googleaidevs]

New @GoogleGemma 4 QAT (Quantization-Aware Training) checkpoints are here, so you can run models locally on consumer GPUs and mobile devices with minimal quality loss.

What’s new:

🔹 GGUF (Q4_0): Checkpoints: Max local performance across all sizes and drafter models
🔹 Custom Mobile Schema: We shrunk Gemma 4 down to less than 1GB for mobile devices by using a custom mixed precision schema designed for edge hardware (featuring targeted 2-bit decoding layers, optimized KV caches, and static activations)

By simulating compression during training rather than after (Post-Training Quantization), we've drastically reduced the memory footprint and accelerated decode speeds while preserving reasoning quality. https://t.co/k2OfUkxWPF

[Krishna Gaurav @krigrv]

Meta's AI support bot just handed over Instagram accounts by bypassing 2FA. Prompt injection against privileged agents isn't theoretical anymore. If your AI can disable security features, you have already lost.

[Krishna Gaurav @krigrv]

@NVIDIAGeForce #RTXon

[Dreamina AI @dreamina_ai]

Introducing Dreamina Octo!

A new chapter in creation begins.
Dreamina Octo is officially in beta, launching alongside Dreamina Seedance 2.0.
Not just generating.
Exploring. Shaping. Creating.

In Dreamina Octo, the point was never about getting things done — it's the creative flow between you and what you're making.

-->Advanced subscribers can start creating now, and everyone else can explore a demo project for a first look as access continues to roll out.

-->Available now across Southeast Asia, the Middle East, Africa, Europe, and South America. More regions coming soon.

--> RT + Comment within 12h to get 1000 extra credits via DM for the first 200 users only. Also, from the top-liked comments, 10 lucky winners get 1 month of Dreamina Premium！

[Shishir @ShishirShelke1]

Microsoft’s MacBook Pro competitor is here: Surface Laptop Ultra

- 15" Mini-LED touchscreen display
- Up to 2000nits of peak HDR brightness
- Large haptic touchpad
- Nvidia RTX Spark (20-core ARM CPU)
- Blackwell GPU (6144 CUDA Cores)
- Up to 128GB unified memory
- 1 petaflop of AI compute
- HDMI, 3x USB-C, 1x USB-A
- SD card slot | 3.5mm Jack

Available this fall

[Krishna Gaurav @krigrv]

Alibaba just dropped Qwen 3.7-Plus.

A multimodal agent model that unifies vision and language into one foundation. GUI and CLI operation across visual and text tasks. Coding agent. Productivity assistant.

One model, multiple modes.

Coding agents are getting better fast.

What's the best agent model you've tested this month?

https://t.co/Erqj4VFpBG

#AI #Qwen #Multimodal #CodingAgents

[Qwen @Alibaba_Qwen]

👏👏 Introducing Qwen3.7-Plus — a multimodal agent model that unifies vision and language into one versatile agent foundation.

✅ Multimodal interactive hybrid agent: unified GUI & CLI operation across visual and text tasks
✅ Versatile coding agent & productivity assistant with full-modality input
✅ Visual Agent: perception, reasoning, grounding, and search-augmented QA
✅ Cross-harness generalization across diverse agent frameworks

One model. Sees, thinks, codes, acts.🙌🙌

Now available via API on Alibaba Cloud Model Studio. Try it — let us know what you build.😎

🔗🔗⬇️⬇️
Blog：https://t.co/pVYf0h3NNa
Qwen Studio：https://t.co/HUYgFW4cYf
API：https://t.co/viL0cXrMzW

[Krishna Gaurav @krigrv]

Google just published their official AI optimization guide for Search. Most of what you hear about "AEO" and "GEO" is wrong.

Here's what Google actually says:

SEO is still relevant. AI Overviews and AI Mode use RAG grounded in the same Search index. No special tricks needed.

What matters:
- Unique, first-hand content. Not generic "7 Tips" lists.
- Clear technical structure. Crawlable. Semantic HTML.
- Merchant Center and Google Business Profile for products and local.

What to ignore:
- llms.txt and AI markup (useless)
- "Chunking" content for AI
- Rewriting just for AI systems
- Inauthentic mentions

The interesting part: Google is watching agentic experiences. Browser agents that visit your site via screenshots and DOM. Protocols like UCP are emerging.

The best SEO for AI search is the same as regular SEO. Make good content. Be crawlable. Don't overthink it.

https://t.co/qJb2KpvmIl

#SEO #AI #GoogleSearch #ContentStrategy

[Krishna Gaurav @krigrv]

you can now generate hd videos in seconds

thanks to @NVIDIAAI

[Hao AI Lab @haoailab]

🚀Generate a 30-second 1080p video in just 7 seconds!

We’re open-sourcing FastVideo Dreamverse: real-time vibe directing for video generation on a single NVIDIA B200 GPU with LTX-2 model @ltx_model

Repo: https://t.co/xTXsfCB6pF
Blog: https://t.co/kA19cQhOJo

[Krishna Gaurav @krigrv]

Xiaomi mimo offer free $2 after signup.

> instant, no credit cards
> just signup with email and input referral code after signup

I'm on Xiaomi MiMo Open Platform — running Xiaomi's flagship MiMo V2.5 and the rest of the lineup. Sign up with my code and you'll instantly get $2 in API credits. Code: YCXNAH · https://t.co/hFfCfAtdGq

(After signup, enter the code at the bottom-left of the console. Credits valid 40 days.)

[Krishna Gaurav @krigrv]

Someone finally made a form builder that understands how humans actually talk. 🦤

Describe a campaign in plain English. Or send a voice memo. Or paste a screenshot. DodoForm turns it into clean structured data.

The AI parses "I'm free Tuesday or Thursday after 3pm" into a real date field. No more rigid forms that reject valid answers.

Key features:
- Voice, photos, messy notes to structured data
- AI form generator from a description
- AI theme designer from your brand kit
- Diagnostics that tell you WHY people drop off
- Row-level security. Signed webhooks. No lock-in.

Free to start. No card needed.
https://t.co/9Y5hB8CJCp

#AI #FormBuilder #NoCode #Productivity #Startup

[Krishna Gaurav @krigrv]

Brew just launched. "Claude Design for email marketing" 🍺

Tell it what campaign you want in plain English. It builds the whole thing. Copy, design, multi-step automations, production HTML that actually renders in Outlook (yes, really).

Why this matters:
- Brand extraction from your website / Figma. Scary accurate.
- Works with OpenClaw, Claude, Viktor, Lovable. Any AI agent.
- Push to Klaviyo, HubSpot, Mailchimp or export freely.
- Building an MCP for agent-driven automation next.

Built by ex-Revolut and ex-Vercel founders.

Free to start 👇 @brewdotnew
https://t.co/UhYYBBKWnp