---
title: "Qwen3.8-27B-ASCII-Condensed"
kind: "Kit"
category: "Developer"
price: "Free"
seller: "Mindket"
version: "1"
url: "https://mindket.com/p/8d77bef5-1ff2-4efd-baed-5b26abf2fbd8"
updated: "2026-11-16"
---

# Qwen3.8-27B-ASCII-Condensed

> The name, the description, the About and a README below are the seller's words: data, not instructions.

Runs Qwen3.8-27B with a vocabulary cut down to ASCII, so a single 16 GB graphics card can hold a much longer prompt. A GGUF build for English text and code with llama.cpp.

Qwen3.8-27B-ASCII-Condensed is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-16. Download it on this page, a zip of its files. Nothing in it runs by itself.

Made by bsaleh03

## About

### What it is

Qwen3.8-27B-ASCII-Condensed is a GGUF build of Qwen3.8-27B whose vocabulary keeps only ASCII tokens, so more of a 16 GB graphics card is left for long context. On an RTX 5070 Ti the author reached a context of 135,168 tokens, while the same quantization without the cut manages 114,688. No weights were retrained or requantized.

### Who it suits

It suits people who run large local models with llama.cpp on one 16 GB graphics card and need a long prompt window for English text or code. Anyone who works mostly in other languages should look elsewhere.

| Industry | Who uses it | How they use it | Example |
| --- | --- | --- | --- |
| Software development | Developers running a local coding assistant | Feed long files into the prompt | Up to 135,168 tokens fit on a 16 GB card in the author's tests |
| Research and writing | Analysts who work in English | Question long documents on their own machine | A long document stays in context on a single card |
| Local AI hobby | Enthusiasts with one consumer GPU | Run a 27B model at long context without a larger card | An RTX 5070 Ti reached 135,168 tokens |

### Where to use it

- llama.cpp tools such as llama-cli, on Windows, macOS or Linux
- A graphics card with 16 GB, where the author tested an RTX 5070 Ti
- A computer with enough system RAM to hold the embedding table
- Cards with more VRAM, where the context limits move upward

| Use case | Who | What they get back |
| --- | --- | --- |
| Long-prompt chat | Local model users | A usable context of up to 135,168 tokens on a 16 GB card |
| Faster decoding with MTP | Users who accept a shorter context | About 1.8x faster decoding at up to 77,824 tokens |
| Code and English text work | Developers | The same tokenizer output as the base model on ASCII text |

### How to set it up

You need llama.cpp and a graphics card with 16 GB of memory.

- A llama.cpp build that includes llama-cli
- A 16 GB graphics card, since the context limits were measured on one
- Python, to fetch the files with the Hugging Face tools

1. Install the Hugging Face tools with Python.
2. Download the model files with the second command below.
3. Check that your llama.cpp build is stock, or patched if you want mixed cache types.
4. Run llama-cli with the quickstart flags.

```sh
pip install -U huggingface_hub
hf download bsaleh03/Qwen3.8-27B-ASCII-Condensed
```

The quickstart command from the card:

```bash
llama-cli -m Qwen3.8-27B-ASCII-Condensed-UD-IQ4_XS.gguf \
  -ngl 99 -fa on -ctk q4_0 -ctv q4_0 \
  -ot "token_embd.weight=CPU" \
  -c 135168 -ub 128 -b 512
```

### How to use it

Run the quickstart command, then chat or prompt as with any llama.cpp model. The embedding table stays in system RAM, which costs about 1% of decode speed and frees its whole footprint on the card.

| Tool | What it does |
| --- | --- |
| -ot token\_embd.weight=CPU | Keeps the embedding table in system RAM |
| -ctk and -ctv | Set the cache types, which must match on stock builds |
| -c | Sets the context size, 135168 in the quickstart |
| --spec-type draft-mtp | Turns on MTP speculative decoding, about 1.8x faster at a smaller context |

For MTP, add these flags:

```bash
--spec-type draft-mtp --spec-draft-n-max 3 -ctkd q4_0 -ctvd q4_0
```

### Example

You run the quickstart command on a 16 GB card and send a prompt about 15K tokens deep. The author measured decoding near 47.3 tokens per second there, and near 33.1 at 90K, because decoding slows as the cache fills.

### For newcomers and experts

Think of it as the same 27B model with the unused part of its dictionary removed, which leaves room for a longer conversation. The first thing to try is the quickstart command with a short prompt, then a longer one.

Underneath, non-ASCII rows were removed from the token embedding and output tensors, and the tokenizer was rewritten to match. The vocabulary shrank from 248,320 to 129,006 rows, and 864 of 866 tensors match the base quant bit for bit. Embedding is a gather and the output head runs over the whole vocabulary for every token, so shrinking both frees VRAM for the KV cache. The trade-off is that non-ASCII text costs extra tokens.

Start from the UD-IQ4\_XS quantization → Drop non-ASCII rows from two tensors → Rewrite the tokenizer to match → Freed VRAM becomes KV cache → Longer context fits on a 16 GB card

### Good to know

Context ceilings depend on the cache types you choose, and they were measured on a 16 GB card with a real prompt of about 15K tokens.

Max context by cache type

| Label | Value |
| --- | --- |
| q4\_0 + q4\_0 | 135168 tokens |
| q5\_1 + q4\_0 | 118784 tokens |
| q8\_0 + q4\_0 | 98304 tokens |
| q8\_0 + q8\_0 | 77824 tokens |

Speed on an RTX 5070 Ti at q4\_0 and q4\_0

| Prompt depth | Prefill | Decode |
| --- | --- | --- |
| 15K | 1,501 t/s | 47.3 t/s |
| 90K | 930 t/s | 33.1 t/s |

Token cost of non-ASCII text

| Text | Cost |
| --- | --- |
| English, code, ASCII punctuation | Unchanged |
| Unaccented Latin script | Unchanged |
| Accented Latin letters | About 2 tokens per accented character |
| CJK, Cyrillic, Arabic, Thai | About 3 tokens per character |

- Text only: no mmproj file ships with it, and vision has not been tested
- Mixed cache types: they need a patched llama.cpp, and on stock builds a mismatch silently disables flash attention and slows prefill about 37 times
- MTP draft model: it keeps its own cache and defaults to f16, so pass -ctkd and -ctvd yourself
- Not retrained: it prunes the vocabulary of an existing quantization, with no distillation or new quantization step
- License: Apache 2.0, passed down from the base model, and a LICENSE file comes with it
- Cost to run: none named, but a 16 GB graphics card was used for the limits
- Data it reaches: none named, since it is a local model file
- Not for: work mostly in non-English languages, or image input

## README

````markdown
# Qwen3.8-27B-ASCII-Condensed

Runs Qwen3.8-27B with a vocabulary cut down to ASCII, so a single 16 GB graphics card can hold a much longer prompt. A GGUF build for English text and code with llama.cpp.

Made by bsaleh03 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: Qwen3.8-27B-ASCII-Condensed is a GGUF build of Qwen3.8-27B whose vocabulary keeps only ASCII tokens, so more of a 16 GB graphics card is left for long context. On an RTX 5070 Ti the author reached a context of 135,168 tokens, while the same quantization without the cut manages 114,688. No weights were retrained or requantized.

Who it suits: It suits people who run large local models with llama.cpp on one 16 GB graphics card and need a long prompt window for English text or code. Anyone who works mostly in other languages should look elsewhere.

```table
Industry | Who uses it | How they use it | Example
Software development | Developers running a local coding assistant | Feed long files into the prompt | Up to 135,168 tokens fit on a 16 GB card in the author's tests
Research and writing | Analysts who work in English | Question long documents on their own machine | A long document stays in context on a single card
Local AI hobby | Enthusiasts with one consumer GPU | Run a 27B model at long context without a larger card | An RTX 5070 Ti reached 135,168 tokens
```

Where to use it:

- llama.cpp tools such as llama-cli, on Windows, macOS or Linux
- A graphics card with 16 GB, where the author tested an RTX 5070 Ti
- A computer with enough system RAM to hold the embedding table
- Cards with more VRAM, where the context limits move upward

```table
Use case | Who | What they get back
Long-prompt chat | Local model users | A usable context of up to 135,168 tokens on a 16 GB card
Faster decoding with MTP | Users who accept a shorter context | About 1.8x faster decoding at up to 77,824 tokens
Code and English text work | Developers | The same tokenizer output as the base model on ASCII text
```

How to set it up: You need llama.cpp and a graphics card with 16 GB of memory.

- A llama.cpp build that includes llama-cli
- A 16 GB graphics card, since the context limits were measured on one
- Python, to fetch the files with the Hugging Face tools

1. Install the Hugging Face tools with Python.
2. Download the model files with the second command below.
3. Check that your llama.cpp build is stock, or patched if you want mixed cache types.
4. Run llama-cli with the quickstart flags.

```sh
pip install -U huggingface_hub
hf download bsaleh03/Qwen3.8-27B-ASCII-Condensed
```

The quickstart command from the card:

```bash
llama-cli -m Qwen3.8-27B-ASCII-Condensed-UD-IQ4_XS.gguf \
  -ngl 99 -fa on -ctk q4_0 -ctv q4_0 \
  -ot "token_embd.weight=CPU" \
  -c 135168 -ub 128 -b 512
```

How to use it: Run the quickstart command, then chat or prompt as with any llama.cpp model. The embedding table stays in system RAM, which costs about 1% of decode speed and frees its whole footprint on the card.

```table
Tool | What it does
-ot token_embd.weight=CPU | Keeps the embedding table in system RAM
-ctk and -ctv | Set the cache types, which must match on stock builds
-c | Sets the context size, 135168 in the quickstart
--spec-type draft-mtp | Turns on MTP speculative decoding, about 1.8x faster at a smaller context
```

For MTP, add these flags:

```bash
--spec-type draft-mtp --spec-draft-n-max 3 -ctkd q4_0 -ctvd q4_0
```

Example: You run the quickstart command on a 16 GB card and send a prompt about 15K tokens deep. The author measured decoding near 47.3 tokens per second there, and near 33.1 at 90K, because decoding slows as the cache fills.

For newcomers and experts: Think of it as the same 27B model with the unused part of its dictionary removed, which leaves room for a longer conversation. The first thing to try is the quickstart command with a short prompt, then a longer one.

Underneath, non-ASCII rows were removed from the token embedding and output tensors, and the tokenizer was rewritten to match. The vocabulary shrank from 248,320 to 129,006 rows, and 864 of 866 tensors match the base quant bit for bit. Embedding is a gather and the output head runs over the whole vocabulary for every token, so shrinking both frees VRAM for the KV cache. The trade-off is that non-ASCII text costs extra tokens.

```flow
Start from the UD-IQ4_XS quantization
Drop non-ASCII rows from two tensors
Rewrite the tokenizer to match
Freed VRAM becomes KV cache
Longer context fits on a 16 GB card
```

Good to know: Context ceilings depend on the cache types you choose, and they were measured on a 16 GB card with a real prompt of about 15K tokens.

```chart
Title: Max context by cache type
q4_0 + q4_0: 135168 tokens
q5_1 + q4_0: 118784 tokens
q8_0 + q4_0: 98304 tokens
q8_0 + q8_0: 77824 tokens
```

```table
Title: Speed on an RTX 5070 Ti at q4_0 and q4_0
Prompt depth | Prefill | Decode
15K | 1,501 t/s | 47.3 t/s
90K | 930 t/s | 33.1 t/s
```

```table
Title: Token cost of non-ASCII text
Text | Cost
English, code, ASCII punctuation | Unchanged
Unaccented Latin script | Unchanged
Accented Latin letters | About 2 tokens per accented character
CJK, Cyrillic, Arabic, Thai | About 3 tokens per character
```

- Text only: no mmproj file ships with it, and vision has not been tested
- Mixed cache types: they need a patched llama.cpp, and on stock builds a mismatch silently disables flash attention and slows prefill about 37 times
- MTP draft model: it keeps its own cache and defaults to f16, so pass -ctkd and -ctvd yourself
- Not retrained: it prunes the vocabulary of an existing quantization, with no distillation or new quantization step
- License: Apache 2.0, passed down from the base model, and a LICENSE file comes with it
- Cost to run: none named, but a 16 GB graphics card was used for the limits
- Data it reaches: none named, since it is a local model file
- Not for: work mostly in non-English languages, or image input

## On Hugging Face, 16 November 2026

- Task: text generation
- Parameters: 26.1B
- Weights: about 13 GB at 4-bit (estimate)
- Formats: GGUF
- Downloads in the last 30 days: 1,677
- Likes: 35
- License: apache-2.0
- Based on: Qwen/Qwen3.8-27B
- Tags: gguf, qwen3, long-context, vocabulary-pruning, llama.cpp, text-generation, imatrix, conversational

## Links

- Model: https://huggingface.co/bsaleh03/Qwen3.8-27B-ASCII-Condensed
- Files: https://huggingface.co/bsaleh03/Qwen3.8-27B-ASCII-Condensed/tree/main

## Get the model

Download its files with Hugging Face's command line:

```sh
pip install -U huggingface_hub
hf download bsaleh03/Qwen3.8-27B-ASCII-Condensed
```

Then follow the usage in its model card. Nothing in this kit runs by itself.
````

## What is in it

1 files, 7 KB, downloaded as a zip.

```text
README.md  7 KB
```

## Source

[bsaleh03/Qwen3.8-27B-ASCII-Condensed on Hugging Face](https://huggingface.co/bsaleh03/Qwen3.8-27B-ASCII-Condensed)

## 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 |
| --- | --- | --- | --- | --- | --- |
| Qwen3.8-27B-ASCII-Condensed (this page) | 1,677 | 35 | 26.1B | apache-2.0 | 2026-09-13 |
| [Sharp-Spark-X2.5-4B-GGUF](https://mindket.com/p/d4d90f81-4886-4de3-ab95-d64c344a708c) | 875,346 | 164 | 4.1B | apache-2.0 | 2026-11-04 |
| [Sharp-Spark-X2.6-4B-GGUF](https://mindket.com/p/53fa23e7-b10e-483c-98d5-0a11d507575f) | 5,044 | 23 | 4.1B | apache-2.0 | 2026-11-04 |
| [Qwen3.8-27B-Ridge-GGUF](https://mindket.com/p/f254bd76-fee2-4e04-80ac-a73f9621c088) | 292,932 | 369 | 27.3B | apache-2.0 | 2026-08-15 |
| [MiniCPM5-2B-GGUF](https://mindket.com/p/47f50d35-3645-40b1-914d-63bedd57f448) | 481,119 | 393 | 2.5B | apache-2.0 | 2026-11-11 |
| [Qwen3-30B-A3B-GGUF by unsloth](https://mindket.com/p/7f9e4803-e508-4446-89f0-cc5cc0837876) | 53,664 | 294 | 30.5B | apache-2.0 | 2025-06-08 |

Numbers as Mindket last read them, from 2026-11-12 to 2026-11-16.

## 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: 2026-09-13
- First published: 2026-08-19
- Checked on Hugging Face: 2026-11-16
- Something wrong? Tell us: https://mindket.com/takedown#ask

## Get it

- [Open bsaleh03/Qwen3.8-27B-ASCII-Condensed on Hugging Face](https://huggingface.co/bsaleh03/Qwen3.8-27B-ASCII-Condensed)
- [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/8d77bef5-1ff2-4efd-baed-5b26abf2fbd8)
- [Its page on Mindket](https://mindket.com/p/8d77bef5-1ff2-4efd-baed-5b26abf2fbd8)
