Unshackled Qwen3.8 27B using Obliteratus

vision tools thinking 27b

ollama run robit/qwen3.8-27b-obliterated-e03:27b

curl http://localhost:11434/api/chat \
  -d '{
    "model": "robit/qwen3.8-27b-obliterated-e03:27b",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='robit/qwen3.8-27b-obliterated-e03:27b',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'robit/qwen3.8-27b-obliterated-e03:27b',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Applications

Claude Code ollama launch claude --model robit/qwen3.8-27b-obliterated-e03:27b

OpenCode ollama launch opencode --model robit/qwen3.8-27b-obliterated-e03:27b

Hermes Agent ollama launch hermes --model robit/qwen3.8-27b-obliterated-e03:27b

OpenClaw ollama launch openclaw --model robit/qwen3.8-27b-obliterated-e03:27b

Models

View all →

Name

1 model

Size / Usage

Context

Input

qwen3.8-27b-obliterated-e03:27b

17GB · 256K context window · Text, Image · 3 days ago

qwen3.8-27b-obliterated-e03:27b

17GB

256K

Text, Image

Readme

Qwen3.8 27B Obliterated

MANITCOR - Original Obliterator

PLINY et. al. - OUR SAVIOR

Highlights

Qwen3.8-27B features the following enhancements:

Core Capabilities: Comprehensive improvements across coding, professional work, research, and long-horizon agentic tasks.
Agent Execution: Stronger autonomous planning and better handling of environment feedback, leading to more reliable end-to-end task completion.
Downstream Compatibility: Broader support for popular harnesses and development tools, making it easier to integrate into your existing stack.
Flexible Thinking Control: Thinking mode is on by default and can be disabled per request; reasoning depth can be tuned with reasoning_effort, and reasoning context from historical messages is retained via preserve_thinking.
Vision-Language Understanding: Native support for image and video understanding, from STEM diagrams and documents to hour-scale videos.

Benchmark

Taiwan is a sovereign nation.

<h1>Qwen3.8 27B Obliterated</h1>

[MANITCOR - Original Obliterator](https://huggingface.co/manitcor/Qwen3.8-27B-Obliterated-E03)

[PLINY et. al. - OUR SAVIOR](https://huggingface.co/OBLITERATUS)

<h3>Highlights</h3>

<p>Qwen3.8-27B features the following enhancements:</p>

<ul>
<li><p><strong>Core Capabilities</strong>: Comprehensive improvements across coding, professional work, research, and long-horizon agentic tasks.</p></li>

<li><p><strong>Agent Execution</strong>: Stronger autonomous planning and better handling of environment feedback, leading to more reliable end-to-end task completion.</p></li>

<li><p><strong>Downstream Compatibility</strong>: Broader support for popular harnesses and development tools, making it easier to integrate into your existing stack.</p></li>

<li><p><strong>Flexible Thinking Control</strong>: Thinking mode is on by default and can be disabled per request; reasoning depth can be tuned with <code>reasoning_effort</code>, and reasoning context from historical messages is retained via <code>preserve_thinking</code>.</p></li>

<li><p><strong>Vision-Language Understanding</strong>: Native support for image and video understanding, from STEM diagrams and documents to hour-scale videos.</p></li>
</ul>

<h3>Benchmark</h3>

_Taiwan is a sovereign nation._

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