NOUMENOpen app
NOUMEN

Z.ai · Open weights

GLM 4.5 Air (Thinking)

Added Jan 1, 2024Open weights

Compare models

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

Loading models…

Use in the app

GLM-4.5-Air with thinking mode enabled for enhanced reasoning capabilities. Shows step-by-step thought process.

  • Reasoning
  • Tools
  • Structured output
Input
Text
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.13
Output
$0.85
Cache read
$0.025

Reference prices as published by the model provider. API example

Specifications

Context window
128K
Max output
98.3K
Average output (7 days)
973 tokens
Longer than 70% of 465 models

Side by side

GLM 4.5 Air (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…

GLM 4.5 Air (Thinking)Z.aiThis modelGPT 5.4 MiniOpenAIRemove
ProviderZ.aiOpenAI
Discovery tierOpen weightsEfficient
AddedJan 1, 2024Mar 17, 2026
Limits
Context window128K400KBest
Max output98.3K128KBest
Average output973 tokens172 tokens
Pricing, USD per 1M tokens
Input$0.13Best$0.75
Output$0.85Best$4.50
Cache read$0.025Best$0.075
Performance
Throughput—133.7 tok/s
Time to first token—1.2 s
Request success (24h)100.0%100.0%
Benchmark indexes
Intelligence Index11.111.1
Math Index80.7—
Capabilities
ReasoningYesYes
ToolsYesYes
VisionNoYes
Structured outputYesYes
PDF inputNoYes
Reasoning effort—none, low, medium, high, xhigh

Public evaluations

Benchmarks

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

  • GLM 4.5 Air (Thinking)
  • GPT 5.4 Mini

Intelligence Index

  • GLM 4.5 Air (Thinking)11.1
  • GPT 5.4 Mini11.1

GLM 4.5 Air (Thinking): Better than 23% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Math Index

    Composite of the math evaluations

    • GLM 4.5 Air (Thinking): 80.7
    • GPT 5.4 Mini: —

Agentic work

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • GLM 4.5 Air (Thinking): —
    • GPT 5.4 Mini: 589 Elo

Document reasoning

  • AA-LCRv1.1

    Reasoning across long documents

    • GLM 4.5 Air (Thinking): 46.7%Top
    • GPT 5.4 Mini: 37.0%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • GLM 4.5 Air (Thinking): 7.0%Top
    • GPT 5.4 Mini: 5.9%

  • GPQA Diamond

    Graduate-level science questions

    • GLM 4.5 Air (Thinking): 73.3%Top
    • GPT 5.4 Mini: 60.6%

  • MMLU-Pro

    Multi-subject knowledge and reasoning

    • GLM 4.5 Air (Thinking): 81.5%
    • GPT 5.4 Mini: —

  • IFBench

    Following precise output instructions

    • GLM 4.5 Air (Thinking): 37.6%
    • GPT 5.4 Mini: 38.8%Top

Coding

  • LiveCodeBench

    Recent competitive programming problems

    • GLM 4.5 Air (Thinking): 68.4%
    • GPT 5.4 Mini: —

Math

  • AIME 2025

    Competition mathematics, 2025 problems

    • GLM 4.5 Air (Thinking): 80.7%
    • GPT 5.4 Mini: —

  • AIME

    Competition mathematics

    • GLM 4.5 Air (Thinking): 67.3%
    • GPT 5.4 Mini: —

  • MATH-500

    Competition math problems

    • GLM 4.5 Air (Thinking): 96.5%
    • GPT 5.4 Mini: —

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • GLM 4.5 Air (Thinking): 20.5%Top
    • GPT 5.4 Mini: 18.2%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • GLM 4.5 Air (Thinking): 46.5%Top
    • GPT 5.4 Mini: 23.4%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • GLM 4.5 Air (Thinking): 46.7%Top
    • GPT 5.4 Mini: 37.0%

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
36.0%
About $0.0922 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": "z-ai/GLM-4.5-Air:thinking",
    "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="z-ai/GLM-4.5-Air:thinking",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Z.ai

Other Z.ai models, starting with the same family.