PolyChat just got a major update!
What’s new in 1.0.4?
🎤 Voice Mode: Speak Your Mind!
We're thrilled to announce Voice Mode, a brand new way to interact with PolyChat. Now you can simply speak your prompts to our model and receive a spoken response back. It's faster, more natural, and completely hands-free.
📲 Quick Share: Search Smarter, Not Harder!
Quick Share makes your search experience smoother than ever. Now, when you're outside the app and spot something interesting, you can simply highlight any text on your screen and share it directly with PolyChat in just two clicks. Instantly search for more information without the hassle of copying and pasting.
https://t.co/CkRrmrQqCu
The hardest part of journaling isn’t writing.
It’s starting when your head feels full.
Quill helps you get thoughts out — by voice or text — and reflect without pressure.
iOS: https://t.co/KXlGXXCBPm
Google Play: https://t.co/iGtqK8vFQn
#Journaling#MentalHealth
Blazing-fast local AI models at the touch of your finger tips.
It works completely offline anywhere, anytime. Get access to unprecedented levels of intelligence running right on your phone.
Check out PolyChat on iOS!
https://t.co/XwnqDDiFQw
@haider1 gpt-4.5 is great at staying in the flow, less likely to overthink, more likely to keep pace with you. Unlike o3 that thinks forever.
that matters when you’re building or writing fast
Many people are raving about the new voice mode on PolyChat!
Haven’t tried it yet? You’re really missing out.
https://t.co/XwnqDDi80Y
#Tryitnow#LLM#AI#Qwen#Qwen3#opensource#voice
@ns123abc Llama is an amazing family of models & what makes it even better is the fact that it’s open source! Even the 1.5 - 3B work are exceptional for summarization, drafting & general knowledge
Too much hype on AGI, not enough focus on extracting value out of today’s models.
We have access to unprecedented intelligence—let’s tap into this potential.
@kimmonismus He did mention that GPT 5 will eliminate the confusion between the model names. Not sure if that means they’ll be merging the capabilities of all models into one. But we’re excited to see what GPT 5 has to offer!
@haider1 These models are usually baked and ready weeks or months before they release them. So I suspect that it’s ready but they’re most likely waiting for the right time to release it to market, just like xAI and others.
@mark_k It may seem that way but with the future of AI so unpredictable. It’s difficult to say things with absolute certainty. Sama lays out his vision for the future.
@nrehiew_@sama understands that AI must transition into a technology that’s more compatible with the daily hardware we use. PolyChat already allows you to run the latest open-source models directly on your phone in case you don’t want to wait around for OpenAI!
https://t.co/XwnqDDiFQw
@ai_for_success Kimi is another perfect example of models getting smaller without deteriorating. 6x smaller than DeepSeek R1 while achieving way better performance. Small open-source model FTW!
🚨Breaking: New DeepSeek-r1 (0528) just tied for #1 in WebDev Arena, matching Claude Opus 4!
More highlights:
💠 #6 Overall on Text Arena
💠 #2 in Coding, #4 in Hard Prompts, #5 in Math category
💠 MIT-licensed, currently the best open model on the leaderboard!
Huge congrats @deepseek_ai on the incredible milestone 🔥