I build things that make me excited about the future. @BottleCapAI: AI Lab focused on efficiency & new architectures. @BeatSaber VR game (acquired by Meta).
Joined July 2012
- Tweets3.4K
- Following821
- Followers18.4K
- Likes5.3K
A fun real-time visualization of my “agent org” in action, powered by our ThinkingCap models running locally on my cluster.
Congrats Tomas! The best yet to come 🚀
Congratulations, Tomáš Mikolov!🏅
Our co-founder and Chief Scientific Officer has received the Medal of Merit from @prezidentpavel🇨🇿 for his work in science and technology.
With more than 200,000 citations and over 20 years in AI, Tomáš is among the world's most accomplished AI scientists. He was among the first members of Google Brain and Facebook AI Research, and his work includes algorithms for scaling neural language models to large datasets.
Now we’re building together the next generation of AI algorithms that increase intelligence while reducing the cost.
Congratulations, Tomáš. Well deserved 👏
Jaroslav Beck retweeted
Mikolov a Beck vylepšili umělou inteligenci Googlu. Je zdarma a výrazně méně náročná
lupa.cz/aktuality/mikolov-a-…
Jaroslav Beck retweeted
One of the most interesting AI labs in Europe is being built out of Prague
The team believe that frontier models drastically overthink, and therefore burn tokens unnecessarily
So @BottleCapAI (named after Europe's greatest innovation) takes an existing open model and trains it to think less while staying just as accurate.
Last week the lab released 3 models:
>> Nvidia’s Nemotron: up to 31% and on average 23% less thinking, with comparable benchmark accuracy
>> Alibaba’s Qwen3.8-Flash-Next: spending 39% fewer thinking tokens on average for about a point of accuracy loss
>> Google’s Gemma 4: their first ThinkingCap on a Google model, with 39% fewer thinking tokens on average for just 0.15pp of accuracy
The impact of what they're building is less token spend and less hardware is required.
VERY COOL
@JaroslavBeck @itsdavidherel
🚨ThinkingCap: Gemma-4-26B-A4B is out!
The most token-efficient Gemma-4-26B-A4B available.
🚀Up to 69% fewer thinking tokens, 38.9% on average
🎯Negligible accuracy loss 0.15pp
🏢ThinkingCap PRO: up to 87.6% fewer thinking tokens, 62% on average - Contact our enterprise[at]bottlecapai[dot]com
📄Blogpost: bottlecapai.com/post/thinkin…
🤗Hugging Face: huggingface.co/bottlecapai/T…
🔗GGUF: huggingface.co/bottlecapai/T…
My wife told me we’re going to a movie themed event next week and that I should find myself an introvert friendly costume… Challenge accepted.
Jaroslav Beck retweeted
The ThinkingCap series keeps growing.
Last week we released:
→ Nemotron 3.5 Lightning: up to 2.3x more requests served at once
→ Qwen3.8-Flash-Next: up to 51% fewer thinking tokens
→ Gemma 4 26B: up to 69% fewer thinking tokens
Have fun!
bottlecapai.com/#products
Jaroslav Beck retweeted
🚨ThinkingCap: Gemma-4-26B-A4B is out!
The most token-efficient Gemma-4-26B-A4B available.
🚀Up to 69% fewer thinking tokens, 38.9% on average
🎯Negligible accuracy loss 0.15pp
🏢ThinkingCap PRO: up to 87.6% fewer thinking tokens, 62% on average - Contact our enterprise[at]bottlecapai[dot]com
📄Blogpost: bottlecapai.com/post/thinkin…
🤗Hugging Face: huggingface.co/bottlecapai/T…
🔗GGUF: huggingface.co/bottlecapai/T…
Third release in 3 days! Let's go! 🚀
🚨ThinkingCap: Gemma-4-26B-A4B is out!
The most token-efficient Gemma-4-26B-A4B available.
🚀Up to 69% fewer thinking tokens, 38.9% on average
🎯Negligible accuracy loss 0.15pp
🏢ThinkingCap PRO: up to 87.6% fewer thinking tokens, 62% on average - Contact our enterprise[at]bottlecapai[dot]com
📄Blogpost: bottlecapai.com/post/thinkin…
🤗Hugging Face: huggingface.co/bottlecapai/T…
🔗GGUF: huggingface.co/bottlecapai/T…
Jaroslav Beck retweeted
We're not done yet with releases❗️
Today we are releasing the most efficient Qwen model in its category: ThinkingCap: Qwen3.8-Flash-Next.
🚨ThinkingCap: Qwen3.8-Flash-Next is out!
The most token-efficient Qwen3.8-Flash-Next available.
🚀Up to 51% fewer thinking tokens, 38.5% on average
🎯Negligible accuracy loss 1.03pp
🏢ThinkingCap PRO: 60% fewer thinking tokens on average - Contact our enterprise[at]bottlecapai [dot]com
📄Blogpost: bottlecapai.com/post/thinkin…
🤗Hugging Face: huggingface.co/bottlecapai/T…
🔗GGUF: huggingface.co/bottlecapai/T…
We're not done yet with releases❗️
Today we are releasing the most efficient Qwen model in its category: ThinkingCap: Qwen3.8-Flash-Next.
🚨ThinkingCap: Qwen3.8-Flash-Next is out!
The most token-efficient Qwen3.8-Flash-Next available.
🚀Up to 51% fewer thinking tokens, 38.5% on average
🎯Negligible accuracy loss 1.03pp
🏢ThinkingCap PRO: 60% fewer thinking tokens on average - Contact our enterprise[at]bottlecapai [dot]com
📄Blogpost: bottlecapai.com/post/thinkin…
🤗Hugging Face: huggingface.co/bottlecapai/T…
🔗GGUF: huggingface.co/bottlecapai/T…
Jaroslav Beck retweeted
ThinkingCap version of NVIDIA Nemotron3.5 Lightning is out. Have fun!
🚨ThinkingCap: NVIDIA Nemotron3.5 Lightning30B A3B is out!
⚡️Up to 2.3x more requests for multi user use compared to original Nemotron version.
🚀 Up to 31%, on average 23% fewer thinking tokens
🎯 Minimal accuracy loss - 0.88pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
ThinkingCap version of NVIDIA Nemotron3.5 Lightning is out. Have fun!
🚨ThinkingCap: NVIDIA Nemotron3.5 Lightning30B A3B is out!
⚡️Up to 2.3x more requests for multi user use compared to original Nemotron version.
🚀 Up to 31%, on average 23% fewer thinking tokens
🎯 Minimal accuracy loss - 0.88pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
Jaroslav Beck retweeted
We made for all interested researchers new open task, the ThinkingCap Test 👉 github.com/BottleCapAI/Think…
Details: bottlecapai.com/careers/natu…
Ok. I am that type of person who instead of cleaning kitchen spends that time by cleaning training data for my humanoid Jarmil to do autonomously. Of course my wife is equally excited as me…
Jaroslav Beck retweeted
🚨 Same model size. Same GPU. ~4.7x more work done!
ThinkingCap: Qwen 3.8 reaches answers with about half the reasoning. On a shared server, that means 2x more users fit in memory at once, and each finishes 2x faster.
Running it solo? ~2x faster answers.
Serving a team? ~4.7x throughput. 🚀
🤗 Hugging Face: huggingface.co/bottlecapai/T…
