turintech/ qwen3.5-0.8b-artemis

56 Downloads Updated 3 hours ago

Text-only Qwen3.5 0.8B GGUF files. Default: q5_k_m. TurinTech models, built with Artemis.

tools thinking

ollama run turintech/qwen3.5-0.8b-artemis

curl http://localhost:11434/api/chat \
  -d '{
    "model": "turintech/qwen3.5-0.8b-artemis",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='turintech/qwen3.5-0.8b-artemis',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'turintech/qwen3.5-0.8b-artemis',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Applications

Claude Code

Claude Code ollama launch claude --model turintech/qwen3.5-0.8b-artemis

OpenCode

OpenCode ollama launch opencode --model turintech/qwen3.5-0.8b-artemis

Hermes Agent

Hermes Agent ollama launch hermes --model turintech/qwen3.5-0.8b-artemis

OpenClaw

OpenClaw ollama launch openclaw --model turintech/qwen3.5-0.8b-artemis

Models

Name

3 models

Size / Usage

Context

Input

qwen3.5-0.8b-artemis:latest

569MB · 256K context window · Text · 17 hours ago

qwen3.5-0.8b-artemis:latest

569MB

256K

Text

qwen3.5-0.8b-artemis:q4_k_m

532MB · 256K context window · Text · 3 hours ago

qwen3.5-0.8b-artemis:q4_k_m

532MB

256K

Text

qwen3.5-0.8b-artemis:q5_k_m

569MB · 256K context window · Text · 3 hours ago

qwen3.5-0.8b-artemis:q5_k_m

569MB

256K

Text

Readme

Qwen3.5 0.8B Artemis

GGUF versions of Qwen3.5 0.8B for text. The vision encoder and the multi-token prediction layer are not included. This is an independent conversion, not affiliated with or endorsed by Alibaba Cloud or the Qwen team.

TurinTech models, built with Artemis

Ollama

ollama run turintech/qwen3.5-0.8b-artemis:q5_k_m

The packages use Ollama’s Qwen3.5 renderer and tool-call parser, the same as Ollama’s own qwen3.5 package, and Qwen’s recommended sampling for text in thinking mode: temperature 1, top_k 20, top_p 0.95, presence penalty 1.5. The context length is not pinned; set it to fit the available memory.

latest selects q5_k_m.

Tag	Hugging Face file	Size
`q5_k_m` (default)	`Qwen3.5-0.8B-Artemis-Q5_K_M.gguf`	0.569 GB
`q4_k_m`	`Qwen3.5-0.8B-Artemis-Q4_K_M.gguf`	0.532 GB

Each tag uses the corresponding Hugging Face file, verified by SHA-256. The comparison measurements used llama.cpp with thinking off, not Ollama.

Measurements and limitations

On the RTX 5090 GSM8K sample, Q5_K_M answered ¹³²⁄₂₅₀ correctly and Q4_K_M ¹³⁷⁄₂₅₀, versus ¹⁴⁵⁄₂₅₀ for BF16. The large tool-calling gains on the earlier panel used tasks that also influenced file selection; the fresh CPU probe did not establish a general tool-use advantage.

Full measurements and test conditions.

Licence

Apache-2.0, as released by Qwen. See the Hugging Face repository for LICENSE and NOTICE.

These files keep improving.

# Qwen3.5 0.8B Artemis

GGUF versions of Qwen3.5 0.8B for text. The vision encoder and the multi-token prediction layer are not included. This is an independent conversion, not affiliated with or endorsed by Alibaba Cloud or the Qwen team.

TurinTech models, built with Artemis

## Ollama

```bash
ollama run turintech/qwen3.5-0.8b-artemis:q5_k_m
```

The packages use Ollama's Qwen3.5 renderer and tool-call parser, the same as Ollama's own `qwen3.5` package, and Qwen's recommended sampling for text in thinking mode: temperature 1, top_k 20, top_p 0.95, presence penalty 1.5. The context length is not pinned; set it to fit the available memory.

`latest` selects `q5_k_m`.

| Tag | Hugging Face file | Size |
| --- | --- | --- |
| `q5_k_m` (default) | `Qwen3.5-0.8B-Artemis-Q5_K_M.gguf` | 0.569 GB |
| `q4_k_m` | `Qwen3.5-0.8B-Artemis-Q4_K_M.gguf` | 0.532 GB |

Each tag uses the corresponding Hugging Face file, verified by SHA-256. The comparison measurements used llama.cpp with thinking off, not Ollama.

## Measurements and limitations

On the RTX 5090 GSM8K sample, Q5_K_M answered 132/250 correctly and Q4_K_M 137/250, versus 145/250 for BF16. The large tool-calling gains on the earlier panel used tasks that also influenced file selection; the fresh CPU probe did not establish a general tool-use advantage.

[Full measurements and test conditions](https://huggingface.co/turintech/Qwen3.5-0.8B-Artemis-GGUF#how-it-was-tested).

## Licence

Apache-2.0, as released by Qwen. See the Hugging Face repository for `LICENSE` and `NOTICE`.

These files keep improving.

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)