← Models
Model APILLM
Qwen·Apache 2.0
Qwen3.8-27B
27B · Dense · Vision-language
Compact dense vision-language model from the Qwen3.8 family — strong coding, agentic execution, and native image/video understanding with flexible thinking control.
Model details
| Developed by | Qwen (Alibaba Cloud) |
| Model family | Qwen3.8 |
| Architecture | Dense VLM (hybrid attention) |
| Parameters | 27B |
| Context length | 262K native (extensible to 1M) |
| Modalities | Text, Image, Video → Text |
| Hidden size | 5120 |
| License | Apache 2.0 |
On Autoloops
| Hardware | H100 |
| API | OpenAI-compatible |
| Region | US West |
| Pricing | $0.40 / $3.00 / $0.15 per 1M tokens (input / output / cached) |
| Model ID | qwen3.8-27b |
Capabilities
- Native vision-language understanding for images and long-form video
- Thinking mode on by default with per-request disable and depth control
- Strong coding, research, and long-horizon agentic task performance
- Open weights under Apache 2.0 for self-host or Autoloops API access
- OpenAI-compatible chat completions via Autoloops
OpenAI-compatible chat completion
from openai import OpenAI
import os
client = OpenAI(
api_key=os.environ["AUTOLOOPS_API_KEY"],
base_url="https://api.autoloops.ai/v1",
)
stream = client.chat.completions.create(
model="qwen3.8-27b",
messages=[{"role": "user", "content": "Tell me a fun fact about cats."}],
stream=True,
)
for chunk in stream:
print(chunk.choices[0].delta.content or "", end="")