NOUMENOpen app
NOUMEN

Qwen · Frontier

Qwen3.8 Max Thinking

Added Aug 3, 2026

Compare models

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

Loading models…

Use in the app

Qwen3.8 Max Thinking enables generation-time reasoning for deeper coding, knowledge work, data analysis, and long-running agent workflows. It supports text, image, video, PDF input, tool calling, structured output, and a near-million-token context window.

  • Reasoning
  • Tools
  • Vision
  • Structured output
  • PDF input
  • Video input
Input
Text, Images, Video
Output
Text
Reasoning effort
low · medium · high · max
Discovery tier
Frontier

Pricing

USD per 1M tokens
Input
$2.00
Output
$6.00
Cache read
$0.25
Cache write
$2.50

Reference prices as published by the model provider. API example

Specifications

Context window
991K
Max output
131.1K

Side by side

Qwen3.8 Max Thinking 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 Max ThinkingQwenThis modelGemini 4 ArgonGoogleRemove
ProviderQwenGoogle
Discovery tierFrontierFrontier
AddedAug 3, 2026Oct 13, 2026
Limits
Context window991K1.05MBest
Max output131.07KBest65.54K
Average output—1.6K tokens
Pricing, USD per 1M tokens
Input$2.00$2.00
Output$6.00Best$10.00
Cache read$0.25$0.10Best
Cache write$2.50$0.19Best
Performance
Throughput—106.6 tok/s
Time to first token—10.0 s
Request success (24h)94.0%98.6%Best
Benchmark indexes
Intelligence Index—52.6
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesYes
PDF inputYesYes
Audio inputNoYes
Video inputYesYes
Reasoning effortlow, medium, high, maxlow, medium, high

Public evaluations

Benchmarks

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

  • Qwen3.8 Max Thinking
  • Gemini 4 Argon

Intelligence Index

  • Qwen3.8 Max Thinking—
  • Gemini 4 Argon52.6
3.8Catalog models by score55.1

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 77.5%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 53.1%

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 3.9%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 1,486 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 1,624 Elo

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 21.8%

  • AA-LCRv1.1

    Reasoning across long documents

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 79.7%

  • MLCR-AA

    Long-context reasoning over medical records

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 13.9%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 57.1%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 57.1%

  • SciCode

    Scientific computing in Python

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 61.8%

Legacy evaluations

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Qwen3.8 Max Thinking: —
    • Gemini 4 Argon: 79.7%

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
96.5%
About $0.31 per 1M input tokens with caching
Request success (24h)
92.4%
Unstable

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 26.4%
  • 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": "qwen/qwen3.8-max:thinking",
    "reasoning_effort": "medium",
    "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="qwen/qwen3.8-max:thinking",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Qwen

Other Qwen models, starting with the same family.