BottleCapAI

🚨Introducing efficiency focused “ThinkingCap” AI model series.  Starting with 2× thinking token reduction on average in Qwen 3.6 27B, with up to 10x faster generation on individual examples. Now available on HuggingFace!

@BottleCapAI's first model is no Top 4 trending on @huggingface 🤯 — behind only GLM-5.2, Hy3 and a Mythos distill. ~350k downloads of the quants in 4 days. The 50% speedup isn't just a claim anymore: verified by multiple community evals (huggingface.co/bottlecapai/Th… and youtu.be/ZTHVEsIEyas). Try it via API on @RequestyAI app.requesty.ai/model-library?…

@codorniou @huggingface 🙌

BottlecAIよりThinkingCap-Qwen3.6-27B-FP8がリリースされた。リリース即300KダウンロードされたThinking長さを半減以下しつつ賢さ維持という特徴を持つモデルの8ビット版だ vLLMで動かすことができるというので、これは是非試したい。さらに明瞭簡潔かもしれないぞ！ huggingface.co/bottlecapai/Th…

@Tono_Ken3 🫶

Ever wished your AI could think ahead and save tokens? Meet ThinkingCap-Qwen3.6-27B, a GGUF model that uses Multi-Token Prediction (MTP) and speculative decoding to generate faster and smarter. It's a game changer for token efficiency.

@HuggingModels 🫶🤗

Best version of Qwen 3.6 27B yet, imho... Huihui-ThinkingCap-Qwen3.6-27B-abliterated-NVFP4. So what makes it different from other 27B quants? * ThinkingCap finetune cuts thinking tokens ~46% with no accuracy loss. Less rambling, same answers. It's abliterated (uncensored) so there is no refusal filter. * NVFP4 W4A4 quantization so it's a mere 20.6 GB, fits on a single 24 GB card. vLLM 0.21+ auto-detects it. * Vision tower kept in bf16 so it is still multimodal! * MTP speculative decoding head kept in bf16. It allows vLLM to draft 3 tokens ahead for faster generation. * Full 128K context at TP=2 (hybrid attention KV is lighter). Many other 27B quants are just the base model compressed. This one stacks shorter reasoning + uncensored + 4-bit + spec decode + vision into one download. Incredible work @support_huihui and @Tono_Ken3

@with_gene2626 🫶

@RequestyAI 🙌

ThinkingCap Qwen3.6-27b is now live on Requesty!

@MiaAI_lab 👀

Testing ThinkingCap-Qwen3.6-27B. I hope to be positively surprised! 🤞 huggingface.co/bottlecapai/Th…

@MalyVitaloshnik Will think about it!

@BottleCapAI Are you planning your official API as well? Even with my unoptimized setup it feels like there could be pretty good margins to support your research :D

No GPU needed. ThinkingCap Qwen3.6-27B (FP8) is now available to everyone through a simple API on Requesty!

Prefer to run it yourself? Weights are free on Hugging Face. 🤗 huggingface.co/bottlecapai/Th… Give it a try and tell us what you build!

Why through Requesty: - OpenAI-compatible API: integrate in 3 lines of code, works with the SDK you already use - Production-ready from day one: 99.99% uptime, automatic failover, EU data residency - Full cost control: real-time analytics, caching, and spending limits built in Launch pricing: $0.40 per million input tokens, $3.00 per million output tokens.

ThinkingCap-Qwen3.6-27B is really good. Much faster than the original Qwen3.6-27B thanks to shorter thinking, while preserving accuracy. Also one of the best evaluation I have seen for a custom Qwen3.6 27B: lot of benchmarks, sigtest, token count, etc. That's all very convincing. I'll publish my own evaluations for this model soon.

@bnjmn_marie 🙌🙏

@vacla_vcerny @JaroslavBeck We are just warming up.

@BottleCapAI @JaroslavBeck By finetuning Chinese models that distill American models.. am I right? ;)

Our name is not a coincidence. Time to speed up!

@RMladek @SeraAureliaX 😉

@SeraAureliaX @BottleCapAI

🤣🤣🤣