Making efficiency-first LLMs & Applications.
- Tweets 168
- Following 4
- Followers 3,480
- Likes 108
Pinned Tweet
🚨ThinkingCap: Qwen3.8-27B is out!
The most token-efficient Qwen3.8 27B available, this time optimized for agentic use.
🚀 Up to 65% fewer thinking tokens, 37% on average
🎯 Minimal accuracy loss
📈 Long-context retrieval actually went up 2.3pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
46
70
782
59,385
BottleCapAI 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…
17
9
136
42,923
🚨 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…
17
9
136
42,923
ThinkingCap: Qwen3.8-27B now runs on a Mac and fits a 32GB machine!
4-bit MLX build - 21 GB, down from 52 GB
Performance is very close to the full precision version.
Vision capabilities and MTP speculative decoding are kept.
huggingface.co/bottlecapai/T…
8
10
154
9,741
And of course here is the correct MLX link! :) huggingface.co/bottlecapai/T…
1
9
747
🚨ThinkingCap: Qwen3.8-27B is out!
The most token-efficient Qwen3.8 27B available, this time optimized for agentic use.
🚀 Up to 65% fewer thinking tokens, 37% on average
🎯 Minimal accuracy loss
📈 Long-context retrieval actually went up 2.3pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
46
70
782
59,385
BottleCapAI retweeted
Actually we were very busy with another big project in works, but I am glad we finally got some time to polish Qwen 3.8 27b with our ThinkingCap. A lot positive feedback so far!🙏
🚨ThinkingCap: Qwen3.8-27B is out!
The most token-efficient Qwen3.8 27B available, this time optimized for agentic use.
🚀 Up to 65% fewer thinking tokens, 37% on average
🎯 Minimal accuracy loss
📈 Long-context retrieval actually went up 2.3pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
2
4
84
6,059
BottleCapAI retweeted
🚨ThinkingCap: Qwen3.8-27B is out!
The most token-efficient Qwen3.8 27B available, this time optimized for agentic use.
🚀 Up to 65% fewer thinking tokens, 37% on average
🎯 Minimal accuracy loss
📈 Long-context retrieval actually went up 2.3pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
46
70
782
59,385
🚨ThinkingCap: Qwen3.8-27B is out!
The most token-efficient Qwen3.8 27B available, this time optimized for agentic use.
🚀 Up to 65% fewer thinking tokens, 37% on average
🎯 Minimal accuracy loss
📈 Long-context retrieval actually went up 2.3pp
📄 Blogpost: bottlecapai.com/post/thinkin…
🤗 Hugging Face: huggingface.co/bottlecapai/T…
🔗 GGUF: huggingface.co/bottlecapai/T…
2
402
If you need help in the future let us know @Alibaba_Qwen 🙌
1
8
1,561
🏢 Enterprise use This release is trained at xhigh effort. For enterprise use, we are ready to fine-tune medium and low effort as well. Contact us at enterprise [at] bottlecapai.com
12
1,979
BottleCapAI retweeted
Official NVFP4 W4A16 version is out!
W4A4 variant is still under consideration, and we’d welcome your feedback. huggingface.co/bottlecapai/T…
6
7
40
3,116