@nizzyabi @orchid_hq I just wanted to get started with orchid but when I had to login it didn't work. Got Server Error. Can you help me with this?
English
Maxi Lieven
466 posts
Maxi Lieven
@maxilvn
17yo Builder | Founder of @TapTalkOfficial
Germany/ Remote شامل ہوئے Temmuz 2025
252 فالونگ62 فالوورز
@argofowl If it gets something wrong or gets stuck just make new session
English
This is so interesting. Imageine somoen selfhosts Kimi and DeepSeek and gets for the price of Gemini 3 flash literally fable level intelligence. Great inovation. Love this
OpenRouter@OpenRouter
Introducing the Fusion API, the smartest compound model in the market. Fusion achieves Fable-level intelligence at half the price. How it works 👇
English
Maxi Lieven ری ٹویٹ کیا
We've open sourced my favorite Devin feature: /handoff
Hand off jobs to cloud Devins from your local machine
Install it as a plugin in Claude Code or Codex or any other coding agent
Close your laptop without pausing your agents 😉
English
This looks pretty promising. Has anyone tried it before?
General Intelligence Company@intelligenceco
Introducing Cofounder's new marketing department. Cofounder sets up your marketing strategy, generates content (video, image, text), connects to channels, runs automatically, and posts on a schedule.
English
Very interesting idea. I can see where this whole agent thing could be heading. Every agent has its own phone, number, ID, ...
Saperly@trysaperly
Introducing a new format for Saperly: Saperly Builders Episode // 01: Pick Up. It’s Your Granny. A behind-the-scenes look at how hard life was before telecom was rebuilt for agents.
English
English
We just shipped PR 1200 - that's 200 more in just 7 days.
That's 20% of all PRs I've ever shipped on @HelloUntangle - in ONE week.
I can't even believe how fast we can ship using a mature full-featured code factory like @DevinAI.
Ryan Carson@ryancarson
PR 1000 landing today. Code Factory is fully functioning. Agents write/review/land 100% of our code now. Don't get me wrong - there's a *lot* of work that I do crafting PRDs, steering conversations, etc, but the majority of the work is done by agents now.
English
@thkostolansky @WisprFlow @aquavoice @ElevenLabs @superwhisper But what is for you most important, speed, accuracy, language, ...?
English
@thkostolansky @WisprFlow @aquavoice @ElevenLabs @superwhisper Imma do a benchmark an I will include Apple then too. Hardest thing is that local models run quite different depending on machine
English
GIF
English
Maxi Lieven ری ٹویٹ کیا
Here’s your monthly reminder that you shouldn’t be prompting coding agents anymore.
You should be designing loops that prompt your agents.
English
I’m in 11th grade and I literally have to learn JavaScript for my computer science exam tomorrow. In the big 2026. 🤣
Ts is lowkey not hard but the school system is so fucking behind
English
@krspy291 you could try TapTalk if you’re on Mac. I built it as a simple local voice input app, so transcription doesn’t depend on Wispr/cloud reliability. curious if it works better with your headphones: taptlk.com
English
is there a better alt to wisprflow at workplace ? just refuses to work well on my headphones
Wispr Flow@WisprFlow
Wispr Flow hasn't been as reliable as it should be lately. We want to be upfront about what happened and what we're doing about it. 🧵
English
@joshearle @WisprFlow I built TapTalk after getting frustrated with reliability too. It’s a simple Mac voice input app where transcription runs locally, so it doesn’t depend on a cloud service being up: taptlk.com
English
Wispr Flow hasn't been as reliable as it should be lately. We want to be upfront about what happened and what we're doing about it. 🧵
English
@pyus13 @WisprFlow I built TapTalk around exactly this concern: voice transcription runs locally on Mac, so your private voice data doesn’t need to be sent to a server. iOS app is coming soon too: taptlk.com
English
Does @WisprFlow send all this private data to the server un-encrypted ? If yes, why not on device AI ?
English
@xadisingh @aradotso How do u make those product videos, they look great
English
@pkaura @WisprFlow Okay, appreciate you. Even that is good feedback, thanks a lot
Imma work on benchmarking TapTalk. Do you have experience in this?
English
@maxilvn @WisprFlow I would love to try but don't want to spend 7 days to figure if this is better - Any benchmarks you have - you will get this objection from most people who want to switch
English
@WisprFlow has become a must have for me but lately there uptime and performance gets a big thumbs down
English
@theayush @WisprFlow Yes it does. The feature is called Refine. Select text, dictate your changes and it what u said will be taken as input prompt to the model and you will get the text out. Speed is same as Wispr or even a bit quicker because we don't do after processing on every dictation
English
@maxilvn @WisprFlow I’d love to, but I’m not sure how it handles updates to the dictation. I use Wispr Flow for the speed, but I also use it to polish what I say, so it corrects pauses, punctuation, and updates. Does TapTalk do that too, or not?
English
What is happening with @WisprFlow
Feels like quality getting degraded, hitting alot of errors. What is happening guys.
English