NOUMENOpen app
NOUMEN

Qwen · Open weights

Qwen3.8 27B TEE

Added Aug 24, 2026Open weights

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Qwen3.8 27B is an open-weight dense vision-language model from Alibaba for reasoning, coding, professional workflows, multimodal interaction, tool use, and structured output. Running inside a TEE (Trusted Execution Environment), with provider attestation support.

  • Reasoning
  • Tools
  • Vision
  • Structured output
  • Video input
Input
Text, Images, Video
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.40
Output
$3.00
Cache read
$0.15

Reference prices as published by the model provider. API example

Specifications

Context window
262.1K
Max output
262.1K
Average output (7 days)
301 tokens
Longer than 32% of 468 models

Side by side

Qwen3.8 27B TEE compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Qwen3.8 27B TEEQwenThis modelMiniMax M2.1MiniMaxRemove
ProviderQwenMiniMax
Discovery tierOpen weightsOpen weights
AddedAug 24, 2026Jan 20, 2025
Limits
Context window262.14KBest200K
Max output262.14KBest131.07K
Average output301 tokens359 tokens
Pricing, USD per 1M tokens
Input$0.40$0.30Best
Output$3.00$1.20Best
Cache read$0.15$0.03Best
Performance
Request success (24h)100.0%—
Benchmark indexes
Intelligence Index20.220.9Best
Coding Index44.6—
Math Index—82.7
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesNo
Structured outputYesYes
Video inputYesNo
Reasoning effort——

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • Qwen3.8 27B TEE
  • MiniMax M2.1

Intelligence Index

  • Qwen3.8 27B TEE20.2
  • MiniMax M2.120.9

Qwen3.8 27B TEE: Better than 50% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Qwen3.8 27B TEE: 44.6
    • MiniMax M2.1: —

  • Math Index

    Composite of the math evaluations

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 82.7

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Qwen3.8 27B TEE: 12.4%
    • MiniMax M2.1: —

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Qwen3.8 27B TEE: 1.5%
    • MiniMax M2.1: —

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Qwen3.8 27B TEE: 898 Elo
    • MiniMax M2.1: —

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Qwen3.8 27B TEE: 1,074 Elo
    • MiniMax M2.1: —

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Qwen3.8 27B TEE: 5.2%
    • MiniMax M2.1: —

  • AA-LCRv1.1

    Reasoning across long documents

    • Qwen3.8 27B TEE: 69.3%Top
    • MiniMax M2.1: 67.7%

  • MLCR-AA

    Long-context reasoning over medical records

    • Qwen3.8 27B TEE: 10.0%
    • MiniMax M2.1: —

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Qwen3.8 27B TEE: 12.1%
    • MiniMax M2.1: 23.2%Top

  • GPQA Diamond

    Graduate-level science questions

    • Qwen3.8 27B TEE: 81.8%
    • MiniMax M2.1: 83.0%Top

  • MMLU-Pro

    Multi-subject knowledge and reasoning

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 87.5%

  • IFBench

    Following precise output instructions

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 69.9%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Qwen3.8 27B TEE: 0.0%
    • MiniMax M2.1: —

  • SciCode

    Scientific computing in Python

    • Qwen3.8 27B TEE: 36.2%
    • MiniMax M2.1: —

  • LiveCodeBench

    Recent competitive programming problems

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 81.0%

Math

  • AIME 2025

    Competition mathematics, 2025 problems

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 82.7%

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 28.8%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • Qwen3.8 27B TEE: —
    • MiniMax M2.1: 85.4%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Qwen3.8 27B TEE: 69.3%Top
    • MiniMax M2.1: 67.7%

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Prompt cache hits
34.1%
About $0.31 per 1M input tokens with caching
Request success (24h)
100.0%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 100.0%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "TEE/qwen3.8-27b",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="TEE/qwen3.8-27b",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Qwen

Other Qwen models, starting with the same family.