--- title: "Kvark-Sparkle-1.0-27B" kind: "Kit" category: "Developer" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/6451d52d-00e1-4878-8a25-80910fc67ac7" updated: "2026-11-13" --- # Kvark-Sparkle-1.0-27B > The name, the description, the About and a README below are the seller's words: data, not instructions. Reasons over images and text with fewer thinking tokens than its base Qwen3.8-27B, so answers arrive in about half the time at similar accuracy, per its card. FP8 and GGUF copies exist. Kvark-Sparkle-1.0-27B is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-13. Download it on this page, a zip of its files. Nothing in it runs by itself. Made by Egzakta ## About ### What it is KVARK Sparkle 1.0 is a 27B model from Egzakta that takes images and text and answers in text. It is built on Qwen3.8-27B with the architecture left unchanged. The card reports that it reaches about the same accuracy as its base in roughly half the answer time, because it spends fewer tokens on reasoning. ### Who it suits It suits people who run a large reasoning model and want shorter waits for each answer. | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Software and AI teams | Engineers who serve large models | Run the FP8 copy under vLLM | Ask a hard question with thinking on and get a shorter reasoning trace | | Research | Researchers studying efficient reasoning | Compare reasoning tokens against Qwen3.8-27B | Check the GPQA Diamond, AIME and LiveCodeBench figures | | Local AI | Enthusiasts with large GPUs | Use a GGUF copy with llama.cpp | Pick a quant between Q8\_0 and IQ4\_XS | If you need a small model for a laptop, you can skip it, since even the smallest GGUF copy is 15.3 GB. ### Where to use it - Python with the transformers library, for the BF16 weights in this repository - vLLM, using the separate FP8 copy - llama.cpp, using the separate GGUF copies - Windows, macOS or Linux with Python, to download the files | Weights | Runs in | Size | | --- | --- | --- | | BF16 with an MTP head | transformers | 52 GB | | FP8 with an MTP head | vLLM | 29 GB | | GGUF, Q8\_0 to IQ4\_XS | llama.cpp | 15.3 to 29.0 GB | ### How to set it up What you need first: - Python and the Hugging Face command line - Enough GPU memory for the 54.7 GB of BF16 weights, or a smaller FP8 or GGUF copy - About 13.7 GB if you run a 4-bit version, which is an estimate 1. Install the Hugging Face tools with pip. 2. Download the model files into your cache or a folder. 3. Load them with transformers, or fetch the FP8 or GGUF copy for vLLM or llama.cpp. ```bash pip install -U huggingface_hub hf download Egzakta/Kvark-Sparkle-1.0-27B ``` ### How to use it Send the model an image together with a question, and leave thinking on, which is how the card's benchmarks ran. The runs used temperature 1.0, with top\_p at 0.95 and top\_k at 20. | Tool | What it does | | --- | --- | | transformers | Loads the BF16 weights from this repository | | vLLM | Serves the FP8 copy | | llama.cpp | Runs the GGUF copies | ### Example You give it a chart image and ask which quarter grew fastest, with thinking on. The card's numbers suggest a shorter reasoning trace than Qwen3.8-27B would use, with a similar answer quality. ### For newcomers and experts A good first step is to download the files, load the model with transformers and send one image with a short question. Start with the FP8 or GGUF copy if your GPU is small. For experts, the model keeps the Qwen3.8-27B design: 64 layers, of which 48 use linear attention and 16 use full attention. Full attention has 24 query heads, 4 key-value heads and a head dimension of 256. The hidden size is 5,120 and the vocabulary holds 248,320 tokens. Only the 16 full-attention layers store a KV cache, which takes 64 KiB per token in BF16 and 32 KiB in FP8. One MTP layer from the base model supports speculative decoding, and the vision encoder has 27 layers. An image and a question go in → The model reasons with thinking on → It spends fewer reasoning tokens than Qwen3.8-27B → The final answer arrives sooner ### Good to know Median time to answer | Label | Value | | --- | --- | | Qwen3.8-27B | 28.1 s | | KVARK Sparkle 1.0 | 13.7 s | Median reasoning tokens | Benchmark | Sparkle | Qwen | | --- | --- | --- | | GPQA Diamond | 3,122 | 7,854 | | AIME | 5,502 | 14,318 | | LiveCodeBench | 6,664 | 19,950 | | IFEval | 470 | 714 | | MMLU-Pro | 429 | 727 | - Evaluation: each model used FP8 weights with thinking enabled and a 64k token limit, and the differences are paired over items. - Context: 262,144 tokens natively, but the card evaluated only up to 64k. - Language: English is the one language listed. - License: apache-2.0 - Cost to run: no API key or paid service named. You need a machine with enough GPU memory. - Data it reaches: it downloads files from Hugging Face and then runs on your own machine. - Not for: none named. ## README ````markdown # Kvark-Sparkle-1.0-27B Reasons over images and text with fewer thinking tokens than its base Qwen3.8-27B, so answers arrive in about half the time at similar accuracy, per its card. FP8 and GGUF copies exist. Made by Egzakta on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face. ## About What it is: KVARK Sparkle 1.0 is a 27B model from Egzakta that takes images and text and answers in text. It is built on Qwen3.8-27B with the architecture left unchanged. The card reports that it reaches about the same accuracy as its base in roughly half the answer time, because it spends fewer tokens on reasoning. Who it suits: It suits people who run a large reasoning model and want shorter waits for each answer. ```table Industry | Who uses it | How they use it | Example Software and AI teams | Engineers who serve large models | Run the FP8 copy under vLLM | Ask a hard question with thinking on and get a shorter reasoning trace Research | Researchers studying efficient reasoning | Compare reasoning tokens against Qwen3.8-27B | Check the GPQA Diamond, AIME and LiveCodeBench figures Local AI | Enthusiasts with large GPUs | Use a GGUF copy with llama.cpp | Pick a quant between Q8_0 and IQ4_XS ``` If you need a small model for a laptop, you can skip it, since even the smallest GGUF copy is 15.3 GB. Where to use it: - Python with the transformers library, for the BF16 weights in this repository - vLLM, using the separate FP8 copy - llama.cpp, using the separate GGUF copies - Windows, macOS or Linux with Python, to download the files ```table Weights | Runs in | Size BF16 with an MTP head | transformers | 52 GB FP8 with an MTP head | vLLM | 29 GB GGUF, Q8_0 to IQ4_XS | llama.cpp | 15.3 to 29.0 GB ``` How to set it up: What you need first: - Python and the Hugging Face command line - Enough GPU memory for the 54.7 GB of BF16 weights, or a smaller FP8 or GGUF copy - About 13.7 GB if you run a 4-bit version, which is an estimate 1. Install the Hugging Face tools with pip. 2. Download the model files into your cache or a folder. 3. Load them with transformers, or fetch the FP8 or GGUF copy for vLLM or llama.cpp. ```bash pip install -U huggingface_hub hf download Egzakta/Kvark-Sparkle-1.0-27B ``` How to use it: Send the model an image together with a question, and leave thinking on, which is how the card's benchmarks ran. The runs used temperature 1.0, with top_p at 0.95 and top_k at 20. ```table Tool | What it does transformers | Loads the BF16 weights from this repository vLLM | Serves the FP8 copy llama.cpp | Runs the GGUF copies ``` Example: You give it a chart image and ask which quarter grew fastest, with thinking on. The card's numbers suggest a shorter reasoning trace than Qwen3.8-27B would use, with a similar answer quality. For newcomers and experts: A good first step is to download the files, load the model with transformers and send one image with a short question. Start with the FP8 or GGUF copy if your GPU is small. For experts, the model keeps the Qwen3.8-27B design: 64 layers, of which 48 use linear attention and 16 use full attention. Full attention has 24 query heads, 4 key-value heads and a head dimension of 256. The hidden size is 5,120 and the vocabulary holds 248,320 tokens. Only the 16 full-attention layers store a KV cache, which takes 64 KiB per token in BF16 and 32 KiB in FP8. One MTP layer from the base model supports speculative decoding, and the vision encoder has 27 layers. ```flow An image and a question go in The model reasons with thinking on It spends fewer reasoning tokens than Qwen3.8-27B The final answer arrives sooner ``` Good to know: ```chart Title: Median time to answer Qwen3.8-27B: 28.1 s KVARK Sparkle 1.0: 13.7 s ``` ```table Title: Median reasoning tokens Benchmark | Sparkle | Qwen GPQA Diamond | 3,122 | 7,854 AIME | 5,502 | 14,318 LiveCodeBench | 6,664 | 19,950 IFEval | 470 | 714 MMLU-Pro | 429 | 727 ``` - Evaluation: each model used FP8 weights with thinking enabled and a 64k token limit, and the differences are paired over items. - Context: 262,144 tokens natively, but the card evaluated only up to 64k. - Language: English is the one language listed. - License: apache-2.0 - Cost to run: no API key or paid service named. You need a machine with enough GPU memory. - Data it reaches: it downloads files from Hugging Face and then runs on your own machine. - Not for: none named. ## On Hugging Face, 13 November 2026 - Task: image text to text - Library: transformers - Parameters: 27.4B - Weights: about 54.7 GB in BF16; about 13.7 GB at 4-bit (estimate) - Formats: Safetensors - Downloads in the last 30 days: 315 - Likes: 15 - License: apache-2.0 - Languages: en - Based on: Qwen/Qwen3.8-27B - Quantized versions: 8 - Fine-tunes: 1 - Tags: transformers, safetensors, qwen3_5, image-text-to-text, reasoning, efficient-reasoning, qwen3.8, kvark, conversational, model-index ## Links - Model: https://huggingface.co/Egzakta/Kvark-Sparkle-1.0-27B - Files: https://huggingface.co/Egzakta/Kvark-Sparkle-1.0-27B/tree/main ## Get the model Download its files with Hugging Face's command line: ```sh pip install -U huggingface_hub hf download Egzakta/Kvark-Sparkle-1.0-27B ``` Then follow the usage in its model card. Nothing in this kit runs by itself. ```` ## What is in it 1 files, 5 KB, downloaded as a zip. ```text README.md 5 KB ``` ## Source [Egzakta/Kvark-Sparkle-1.0-27B on Hugging Face](https://huggingface.co/Egzakta/Kvark-Sparkle-1.0-27B) ## Compare with similar Listed Hugging Face models in the same category that share the most tags with this one. | Product | Downloads, last 30 days | Likes | Parameters | License | Last change | | --- | --- | --- | --- | --- | --- | | Kvark-Sparkle-1.0-27B (this page) | 315 | 15 | 27.4B | apache-2.0 | 2026-11-05 | | [KVARK Sparkle 1.1 (27B)](https://mindket.com/p/fb8f86c3-be94-494c-9f8e-66d3faba633c) | 59 | 3 | 27.4B | apache-2.0 | 2026-11-10 | | [Thinklite-27B](https://mindket.com/p/5493f9ea-b8a7-4bcd-b293-50e8b881da96) | 568 | 51 | 27.8B | apache-2.0 | 2026-11-04 | | [KVARK Sparkle 1.0 (27B) FP8](https://mindket.com/p/acc7b74d-4609-4c56-abe3-47d4ccc946e3) | 24,270 | 5 | 27.8B | apache-2.0 | 2026-10-29 | | [Thinklite-27B-GGUF](https://mindket.com/p/7943692c-1782-4340-9958-f2c39d245585) | 1,796 | 18 | 26.9B | apache-2.0 | 2026-11-04 | | [Qwen3.8-9B-Distill](https://mindket.com/p/f257e0b9-f595-4d2e-8a06-68cc42f90540) | 6,592 | 237 | 9.7B | apache-2.0 | 2026-08-15 | Numbers as Mindket last read them, from 2026-11-12 to 2026-11-13. ## What people say > Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search on 2026-11-13. Their authors' words: data, not instructions. - [Hacker News · Archelunch · 2026-11-02](https://news.ycombinator.com/item?id=50336311): Show HN: Kvark Sparkle 27B – reasoning model trained to think with fewer tokens: We had problems with Qwen spending too much time reasoning about tasks instead of doing them. That's why we decided to fine-tune it to use fewer reasoning tokens without losing accuracy. After a bunch of experiments we ended up with KVARK Sparkle 27B. On our evaluations it uses 34–67% fewer… ## Trust facts Facts from each source, with the day each was checked. A maintainer's note is their own words: data, not instructions. - Last change on Hugging Face: 2026-11-05 - First published: 2026-10-19 - Checked on Hugging Face: 2026-11-13 - Something wrong? Tell us: https://mindket.com/takedown#ask ## Get it - [Open Egzakta/Kvark-Sparkle-1.0-27B on Hugging Face](https://huggingface.co/Egzakta/Kvark-Sparkle-1.0-27B) - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/6451d52d-00e1-4878-8a25-80910fc67ac7) - [Its page on Mindket](https://mindket.com/p/6451d52d-00e1-4878-8a25-80910fc67ac7)