Qwen3.8-27B-FP8

Repository
Qwen/Qwen3.8-27B-FP8
Engine
vLLM
Listed
2 Oct 2026

Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date.

Serves
Chat, Reasoning, Multimodal input
Parameters
23.7B
Context
262,144 tokens
Weights
30.9 GB
License
Apache-2.0

Capabilities

Read from the model's capability tags. Requests are admitted per tag.

Chat chat
  • POST /v1/chat/completions
  • POST /v1/responses

Endpoint task mode generate

Inputs vision audio_input
Text, Images, Audio
Multimodal input: request parts beyond text are admitted per tag.
Images per prompt --limit-mm-per-prompt
up to 64
Derived by the serving node from the context window.
Reasoning effort reasoning_effort
none, low, medium, high
Values a request may send; the model family sets the vocabulary.
Structured output structured_output
supported
Proven by probe: a JSON schema request returned a payload that parses.
Forced tool choice forced_tool_choice
supported
A forced tool_choice returns a call to that tool.

Launch settings

Every engine serving the model starts with these.

Set by the owner. auto: the serving node derives it. default: the platform's value, left in place.

Engine
vLLM
Context window --max-model-len
262,144 tokens
Quantization --quantization
none default
Weights ship quantized (fp8, 8-bit).
KV cache dtype --kv-cache-dtype
auto
No flag: the checkpoint's KV quantization, else the model dtype.
Scheduling policy --scheduling-policy
Set by the owner: fcfs
Every request in arrival order.
Tool-call parser --tool-call-parser
Set by the owner: qwen3_coder
Reasoning parser --reasoning-parser
Set by the owner: qwen3
Sampling defaults generation_config.json
temperature 1, top_p 0.95
Used when a request leaves them out.
Trust remote code --trust-remote-code
on default
Python shipped in the bundle may run at load (the default).

Engine tuning

Every setting auto: the serving node derives each one.

Set by the owner. auto: the serving node derives it.

Mixture of experts

Expert parallel --enable-expert-parallel
auto
Expert load balancing --enable-eplb
auto
All-to-all backend --all2all-backend
auto

Memory and batching

GPU memory utilization --gpu-memory-utilization
auto
Max sequences --max-num-seqs
auto
Max batched tokens --max-num-batched-tokens
auto
Long-prefill threshold --long-prefill-token-threshold
auto

Execution

Enforce eager --enforce-eager
auto
Prefix caching --enable-prefix-caching
auto
Async scheduling --no-async-scheduling
auto
Custom all-reduce --disable-custom-all-reduce
auto

Speculative decoding

Draft tokens checked by the model in one pass.

Speculative decoding
Set by the owner: on for every endpoint
Method
Set by the owner: dflash
A DFlash block drafter trained for the target.
Speculative tokens num_speculative_tokens
Set by the owner: 5
Draft model
Set by the owner: Qwen3.8-27B-DFlash2

Architecture

Read from the model's files at upload.

Architecture
qwen3_5
Parameters
23,682,744,320
Layers num_hidden_layers
64
Hidden size hidden_size
5,120
Attention heads num_attention_heads
24 (4 KV)
Head dim head_dim
256
Vocabulary vocab_size
248,320
Native dtype torch_dtype
bfloat16
Weights on disk
30.9 GB, safetensors

Call it

Create an endpoint for the model, then send it a request in the OpenAI shape.

Shell
# Create an endpoint for this model, then provision it. # --tier-id is the service tier to run on; pick one at /endpoints/new. erectl endpoints create \ --name "My endpoint" \ --slug my-endpoint \ --model-id E71AE9D5-85D2-4E3E-B22C-2B18917BE6B2 \ --tier-id <tier-id> # provision takes the endpoint UUID returned by create erectl endpoints provision <endpoint-id> # Call it. Requests are routed by /<project-id>/<endpoint-slug>. curl https://api.erebine.ai/proj_ABC123/my-endpoint/v1/chat/completions \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $EREBINE_API_KEY" \ -d '{ "model": "Qwen3.8-27B-FP8", "messages": [{"role": "user", "content": "Hello, world!"}], "stream": true }'

Model card

As published with Qwen/Qwen3.8-27B-FP8. Images are left out; its HTML tables are set as tables.

---
base_model:

  • Qwen/Qwen3.8-27B

library_name: transformers
license: apache-2.0
pipeline_tag: image-text-to-text
---

Qwen3.8-27B-FP8

Note. This repository contains FP8-quantized model weights and configuration files for the post-trained model in the Hugging Face Transformers format.
These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, TokenSpeed, etc.
The quantization method is fine-grained fp8 quantization with block size of 128, and its performance metrics are nearly identical to those of the original model.
Tip. For users seeking managed, scalable inference without infrastructure maintenance, the official Qwen API service is provided by Qwen Cloud.
In particular, Qwen3.8-27B will be available as a hosted version with more production features, e.g., 1M context length by default, official built-in tools. For more information, please refer to the Qwen3.8-27B Overview. The service is coming soon. Stay tuned for updates.

Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability.

Qwen3.8 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.

Model Overview

  • Type: Causal Language Model with Vision Encoder
  • Training Stage: Pre-training & Post-training
  • Language Model
  • Number of Parameters: 27B
  • Hidden Dimension: 5120
  • Token Embedding: 248,320 (Padded)
  • Number of Layers: 64
  • Hidden Layout: 16 × (3 × (Gated DeltaNet → FFN) → 1 × (Gated Attention → FFN))
  • Gated DeltaNet:
  • Number of Linear Attention Heads: 48 for V and 16 for QK
  • Head Dimension: 128
  • Gated Attention:
  • Number of Attention Heads: 24 for Q and 4 for KV
  • Head Dimension: 256
  • Rotary Position Embedding Dimension: 64
  • Feed Forward Network:
  • Intermediate Dimension: 17,408
  • LM Output: 248,320 (Padded)
  • MTP (Multi-Token Prediction): trained with multiple steps
  • Context Length: 262,144 natively and extensible up to 1,000,000 tokens.

Benchmark Results

Text Performance
Qwen3.8-27BQwen3.6-27BQwen3.7-PlusMuse Glimmer-30BOpus4.6 Max
Coding
Agentic terminal coding Terminal Bench 2.1 (Terminus)73.063.464.051.778.2
Agentic coding SWE-bench Pro61.753.557.651.253.4
Repo-level code generation NL2Repo-Bench42.336.241.1--47.6
Agentic coding DeepSWE 1.142.213.314.2----
Software engineering QwenSWEBench79.049.359.2--63.8
Agent
Long-horizon office work CoWorkBench70.761.065.1--68.2
Professional job tasks JobBench33.421.827.6----
Frontier agentic tasks Agents' Last ExamPass@1 20.4 Score 42.9Pass@1 10.6 Score 27.3Pass@1 13.2 Score 33.6----
General
Instruction following IFBench79.569.179.177.062.5
Scientific reasoning GPQA Diamond89.287.890.383.591.3
Multidisciplinary reasoning HLE30.824.034.722.040.0
Competitive coding LiveCodeBench v690.383.989.6--88.8

SWE-bench Pro: Except for Opus4.6 Max, which uses the officially reported score, all models are evaluated with the Claude Code harness at temp=1.0, top_p=0.95, and a 256K context window. Problematic tasks were corrected, and all baseline models were re-evaluated on the refined benchmark.
NL2Repo-Bench: Evaluated with the Claude Code harness. To prevent reward hacking, we disable Bash commands that attempt to access the specific repository, such as pip download, pip install, and git clone.
DeepSWE 1.1: Evaluated with the Claude Code harness at temp=1.0, top_p=0.95, and a 256K context window.
QwenSWEBench: In-house coding benchmark for evaluating models' software engineering capabilities. Evaluated with the Claude Code harness. Reporting avg@3 with an 8-hour timeout, max_tokens=32,768, temperature=1.0, and a 256K context window.
CoWorkBench: In-house cowork benchmark for evaluating long-horizon tasks across computer science, finance, law, medical, and other productivity domains.
HLE: Judged by GPT-4o.
The best result in each row is shown in bold.
Empty cells (--) indicate that results are not yet available or not applicable.

VL Performance
Qwen3.8-27BQwen3.6-27BQwen3.7-PlusMuse Glimmer-30BOpus4.6 Max
Agentic Multimodal Intelligence
Computer use OSWorld-Verified84.363.973.365.972.7
Browser use WebArena-Verified64.848.855.3----
Mobile use AndroidWorld81.970.381.0--62.0
Application recreation RecreationBench47.129.830.2----
Multimodal tool use ClawEval-MMPass@3 57.4 Average 56.9Pass@3 42.6 Average 50.4Pass@3 57.4 Average 60.1--Pass@3 52.5 Average 54.7
Multimodal software engineering SWE-MM38.625.730.0--27.1
Visual web development Vision2Web62.945.042.1----
General Multimodal Intelligence
Visual math problem solving MathVisionWithout CI 90.0 With CI 94.6Without CI 85.1Without CI 90.3--Without CI 65.5
General visual reasoning BabyVisionWithout CI 65.7 With CI 85.6Without CI 28.9Without CI 64.7 With CI 70.4--Without CI 12.6
Scientific chart analysis CharXiv (RQ)Without CI 83.7 With CI 90.2Without CI 78.4Without CI 85.8 With CI 85.978.8Without CI 66.0
Document intelligence OmniDocBench 1.591.189.491.475.886.6
Real-world perception RealWorldQA85.984.186.9--73.9
Embodied intelligence ERQA65.562.569.8--40.8

MathVision, BabyVision, and CharXiv (RQ): Where both settings are available, cells report “Without CI” and “With CI” separately; otherwise, only the available setting is shown. A small number of incorrect ground-truth annotations in MathVision and CharXiv (RQ) were corrected following manual verification, and all reported scores on those benchmarks were computed using the corrected annotations.
MathVision: Qwen3.8-27B is evaluated using the fixed prompt: “Please reason step by step, and put your final answer within \boxed{}.” For the remaining models, we report the higher score from two prompt variants—one with and one without the \boxed{} formatting requirement.
WebArena-Verified: Scores are computed with the official WebArena-Verified grader under the OSWorld scaffold.
RecreationBench: An in-house, long-horizon application-recreation benchmark designed to evaluate hybrid-agent capabilities across five platforms: desktop (Ubuntu, macOS, and Windows), mobile (Android), and the web.
ClawEval-MM: Scores are reported as “Pass@3 / average score.” Pass@3 is the percentage of tasks passed in at least one of three trials; the average score is the mean benchmark score across the three trials.
Vision2Web: Scores are averaged across the frontend, webpage, and website categories. Evaluations use the Claude Code harness and are judged by gpt-5.4-2026-03-05.
SWE-MM: Scores are evaluated on the Claude Code harness using the public dev split of SWE-bench Multimodal, with the modifications described in Appendix 8.3 of the Claude Opus 4.7 system card.
Empty cells (--) indicate that results are not yet available or not applicable.

Quickstart

For streamlined integration, we recommend using Qwen3.8 via APIs.

Serving Qwen3.8
Important. Inference efficiency and throughput vary significantly across frameworks. We recommend using the latest framework versions to ensure optimal performance and compatibility. For production workloads or high-throughput scenarios, dedicated serving engines such as SGLang, vLLM, or TokenSpeed are recommended.

Qwen3.8 can be deployed with popular inference frameworks, e.g.:

API Usage
Important. Qwen3.8 models operate in thinking mode by default, generating thinking content signified by \n...\n\n before producing the final response. To disable thinking content and obtain a direct response, refer to the examples here.
Tip. We recommend using the following sets of sampling parameters for generation:
  • Thinking Mode: temperature=1.0, top_p=0.95, top_k=20, min_p=0.0, presence_penalty=0.0, repetition_penalty=1.0
  • Instruct (or non-thinking) mode: temperature=0.7, top_p=0.80, top_k=20, min_p=0.0, presence_penalty=1.5, repetition_penalty=1.0

Please note that the support for sampling parameters varies according to inference frameworks.

Qwen3.8 comes with official support for reasoning_effort, which can be used to adjust reasoning depth and control cost:

  • xhigh (default): for complex tasks demanding thorough analysis
  • medium: balancing accuracy and speed
  • low: efficient reasoning optimizing for speed and cost

In addition, preserve_thinking is enabled by default for all workloads for the best out-of-the-box experience. To disable preserved thinking, refer to the examples here.

Tip. In multi-turn agentic tasks, lower reasoning effort does not always reduce overall task completion time. Although it may produce faster per-turn responses, it can also lead to insufficient analysis, more failures, and repeated retries, which may increase total latency and token consumption.
Chat Completions API

The Chat Completions API can be used with most inference frameworks, as well as Qwen Cloud.
Before starting, make sure the OpenAI Python SDK is installed and the API key and the API base URL are configured, e.g.:

pip install -U openai

# Set the following accordingly
export OPENAI_BASE_URL='your-base-url'
export OPENAI_API_KEY='your-api-key'
Text-Only Input
from openai import OpenAI
# Configured by environment variables
client = OpenAI()

messages = [{"role": "user", "content": "Write a Python function to merge two sorted linked lists."}]

completion = client.chat.completions.create(
    model="Qwen/Qwen3.8-27B-FP8",
    messages=messages,
    extra_body={
        "chat_template_kwargs": {
            "enable_thinking": True,  # on by default
            "preserve_thinking": True, # on by default
        },
    },
    reasoning_effort="xhigh",  # xhigh by default; supported levels are xhigh, medium, and low
    stream=True,
    stream_options={"include_usage": True},
)

reasoning_content = ""
answer_content = ""
is_answering = False
print("\n" + "=" * 20 + "Reasoning" + "=" * 20 + "\n")

for chunk in completion:
    if not chunk.choices:
        print("\nUsage:")
        print(chunk.usage)
        continue

    delta = chunk.choices[0].delta

    if hasattr(delta, "reasoning_content") and delta.reasoning_content is not None:
        if not is_answering:
            print(delta.reasoning_content, end="", flush=True)
        reasoning_content += delta.reasoning_content
    elif hasattr(delta, "reasoning") and delta.reasoning is not None:
        if not is_answering:
            print(delta.reasoning, end="", flush=True)
        reasoning_content += delta.reasoning

    if hasattr(delta, "content") and delta.content:
        if not is_answering:
            print("\n" + "=" * 20 + "Answer" + "=" * 20 + "\n")
            is_answering = True
        print(delta.content, end="", flush=True)
        answer_content += delta.content

messages.append({
    "role": "assistant",
    "content": answer_content,
    "reasoning_content": reasoning_content,
    "reasoning": reasoning_content,
})
Image Input
from openai import OpenAI
# Configured by environment variables
client = OpenAI()

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image_url",
                "image_url": {
                    "url": "https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3.5/demo/CI_Demo/mathv-1327.jpg"
                }
            },
            {
                "type": "text",
                "text": "The centres of the four illustrated circles are in the corners of the square. The two big circles touch each other and also the two little circles. With which factor do you have to multiply the radii of the little circles to obtain the radius of the big circles?\nChoices:\n(A) $\\frac{2}{9}$\n(B) $\\sqrt{5}$\n(C) $0.8 \\cdot \\pi$\n(D) 2.5\n(E) $1+\\sqrt{2}$"
            }
        ]
    }
]

chat_response = client.chat.completions.create(
    model="Qwen/Qwen3.8-27B-FP8",
    messages=messages,
)
print("Chat response:", chat_response)
Video Input

INLINECODE21z9844948A, # video frame sampling can be configured via extra_body (e.g., by setting fps). # This feature is currently supported only in vLLM. # # By default, fps=2 and do_sample_frames=True. # With do_sample_frames=True, you can customize the fps value to set your desired video sampling rate. # chat_response = client.chat.completions.create( # model="Qwen/Qwen3.8-27B-FP8", # messages=messages, # extra_body={ # "mm_processor_kwargs": {"fps": 2, "do_sample_frames": True}, # }, # ) print("Chat response:", chat_response)


##### Instruct (or Non-Thinking) Mode

Qwen3.8-27B will think by default before responding.
You can obtain a direct response from the model without thinking by configuring the API parameters. 
For example,
python from openai import OpenAI # Configured by environment variables client = OpenAI() messages = [ { "role": "user", "content": [ { "type": "image_url", "image_url": { "url": "https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3.5/demo/RealWorld/RealWorld-04.png" } }, { "type": "text", "text": "Where is this?" } ] } ] chat_response = client.chat.completions.create( model="Qwen/Qwen3.8-27B-FP8", messages=messages, temperature=0.7, top_p=0.8, presence_penalty=1.5, extra_body={ "top_k": 20, "chat_template_kwargs": {"enable_thinking": False}, }, ) print("Chat response:", chat_response) INLINECODE29z1E4D23DEmodel, please use "enable_thinking": False instead of "chat_template_kwargs": {"enable_thinking": False}.

Disable Preserved Thinking

By default, Qwen3.8 retains thinking blocks from all historical messages, maintaining a complete reasoning trace across the conversation. This behavior, known as preserved thinking, ensures full context continuity and is especially beneficial for agent scenarios where decision consistency and reduced redundant reasoning are critical. It also improves KV cache utilization, optimizing inference efficiency in both thinking and non-thinking modes.

If you prefer to retain only the thinking blocks from the latest user message, you can disable this behavior by setting preserve_thinking to False:

from openai import OpenAI

# Configured by environment variables
client = OpenAI()
messages = [...]
chat_response = client.chat.completions.create(
    model="Qwen/Qwen3.8-27B-FP8",
    messages=messages,
    extra_body={
        "chat_template_kwargs": {"preserve_thinking": False},
    },
)
print("Chat response:", chat_response)
Note. If you are using APIs from Qwen Cloud, in addition to changing model, please use "preserve_thinking": False directly instead of wrapping it in chat_template_kwargs.

Best Practices

To achieve optimal performance, we recommend the following settings:

  1. Sampling Parameters: We suggest using the following sets of sampling parameters:
  • Thinking Mode: temperature=1.0, top_p=0.95, top_k=20, min_p=0.0, presence_penalty=0.0, repetition_penalty=1.0
  • Instruct (or non-thinking) mode: temperature=0.7, top_p=0.80, top_k=20, min_p=0.0, presence_penalty=1.5, repetition_penalty=1.0

For supported frameworks, you can adjust the presence_penalty parameter between 0 and 2 to reduce endless repetition. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.

  1. Adequate Output Length: To optimize performance on agentic tasks, we recommend allocating sufficient output length to allow the model to generate detailed and comprehensive responses. For frameworks that support separate token limits for internal reasoning and final outputs, we suggest the following configuration within the 1M context length:
  • Reasoning Content: Set the maximum output length to 262,144 tokens.
  • Final Response: Set the maximum output length to 131,072 tokens.

These settings provide the necessary capacity for complex reasoning while ensuring ample space for high-quality final deliverables.

  1. Processing Ultra-Long Texts: Qwen3.8-27B natively supports context lengths of up to 262,144 tokens. For long-horizon tasks where the total length (including both input and output) exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively, e.g., YaRN.

YaRN is currently supported by several inference frameworks, e.g., vLLM, SGLang, and TokenSpeed.
In general, there are two approaches to enabling YaRN for supported frameworks:

  • Modifying the model configuration file:

In the config.json file, change the rope_parameters fields in text_config to:
INLINECODE55z859CAD5E

  • Passing command line arguments:

For vLLM, you can use
INLINECODE58zF9A29D7C

For SGLang, you can use
INLINECODE61z4D697571

For TokenSpeed, you can use
INLINECODE64zA9E87EE8

> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, potentially impacting performance on shorter texts.
> We advise modifying the rope_parameters configuration only when processing long contexts is required.
> It is also recommended to modify the factor as needed. For example, if the typical context length for your application is 524,288 tokens, it would be better to set factor as 2.0.

  1. Long Video Understanding: To optimize inference efficiency for plain text and images, the size parameter in the released video_preprocessor_config.json is conservatively configured. It is recommended to set the longest_edge parameter in the video_preprocessor_config file to 469,762,048 (corresponding to 224k video tokens) to enable higher frame-rate sampling for hour-scale videos and thereby achieve superior performance. For example,
    {"longest_edge": 469762048, "shortest_edge": 4096}
    CODEBLOCK5z6ECCD421bibtex
@misc{qwen38,
    title = {{Qwen3.8-Max}: A New Bar for Coding and Cowork},
    url = {https://qwen.ai/blog?id=qwen3.8},
    author = {{Qwen Team}},
    month = {August},
    year = {2026}
}