NOUMENOpen app
NOUMEN

Z.ai · Open weights

GLM 5.3 Thinking

Added Aug 14, 2026Open weights

Compare models

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

Loading models…

Use in the app

GLM-5.3 with higher reasoning enabled for harder long-horizon coding, autonomous agent workflows, and complex engineering tasks.

  • Reasoning
  • Tools
  • Structured output
Input
Text
Output
Text
Reasoning effort
high · max
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.60
Output
$1.10
Cache read
$0.20

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
131.1K
Average output (7 days)
2.3K tokens
Longer than 93% of 468 models

Side by side

GLM 5.3 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 5.3 ThinkingZ.aiThis modelDeepSeek V4.1 Flash 1017 ThinkingDeepSeekRemove
ProviderZ.aiDeepSeek
Discovery tierOpen weightsOpen weights
AddedAug 14, 2026Oct 17, 2026
Limits
Context window1.05M1.05M
Max output131.07K393.22KBest
Average output2.3K tokens1.2K tokens
Pricing, USD per 1M tokens
Input$0.60$0.10Best
Output$1.10$0.60Best
Cache read$0.20$0.003Best
Performance
Throughput123.3 tok/s214.9 tok/sBest
Time to first token7.7 s2.4 sBest
Request success (24h)99.7%99.8%Best
Benchmark indexes
Intelligence Index44.845.1Best
Coding Index74.8—
Capabilities
ReasoningYesYes
ToolsYesYes
VisionNoYes
Structured outputYesYes
Reasoning efforthigh, maxnone, low, high, max

Public evaluations

Benchmarks

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

  • GLM 5.3 Thinking
  • DeepSeek V4.1 Flash 1017 Thinking

Intelligence Index

  • GLM 5.3 Thinking44.8
  • DeepSeek V4.1 Flash 1017 Thinking45.1

GLM 5.3 Thinking tested at max reasoning effort

GLM 5.3 Thinking: Better than 90% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • GLM 5.3 Thinking: 74.8
    • DeepSeek V4.1 Flash 1017 Thinking: —

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • GLM 5.3 Thinking: 62.2%
    • DeepSeek V4.1 Flash 1017 Thinking: 66.6%Top

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • GLM 5.3 Thinking: 29.4%
    • DeepSeek V4.1 Flash 1017 Thinking: 36.7%Top

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • GLM 5.3 Thinking: 0.3%
    • DeepSeek V4.1 Flash 1017 Thinking: —

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • GLM 5.3 Thinking: 1,505 Elo
    • DeepSeek V4.1 Flash 1017 Thinking: 1,570 EloTop

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • GLM 5.3 Thinking: 1,649 Elo
    • DeepSeek V4.1 Flash 1017 Thinking: 1,746 EloTop

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • GLM 5.3 Thinking: 11.2%
    • DeepSeek V4.1 Flash 1017 Thinking: 17.2%Top

  • AA-LCRv1.1

    Reasoning across long documents

    • GLM 5.3 Thinking: 79.7%
    • DeepSeek V4.1 Flash 1017 Thinking: 82.7%Top

  • MLCR-AA

    Long-context reasoning over medical records

    • GLM 5.3 Thinking: 17.2%
    • DeepSeek V4.1 Flash 1017 Thinking: 26.7%Top

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • GLM 5.3 Thinking: 42.3%
    • DeepSeek V4.1 Flash 1017 Thinking: 43.3%Top

  • GPQA Diamond

    Graduate-level science questions

    • GLM 5.3 Thinking: 91.7%
    • DeepSeek V4.1 Flash 1017 Thinking: —

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • GLM 5.3 Thinking: 41.9%Top
    • DeepSeek V4.1 Flash 1017 Thinking: 33.8%

  • SciCode

    Scientific computing in Python

    • GLM 5.3 Thinking: 59.0%Top
    • DeepSeek V4.1 Flash 1017 Thinking: 55.9%

Legacy evaluations

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • GLM 5.3 Thinking: 79.7%
    • DeepSeek V4.1 Flash 1017 Thinking: 82.7%Top

Scores from Artificial Analysis, updated Oct 26, 2026. GLM 5.3 Thinking was tested at max reasoning effort; compare scores measured at the same effort.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.

Throughput
123.3 tok/s
Faster than 70% of 254 models
Time to first token
7.7 s
Quicker than 18% of 254 models
Prompt cache hits
46.7%
About $0.41 per 1M input tokens with caching
Request success (24h)
99.7%
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 98.8%
  • 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-5.3:thinking",
    "reasoning_effort": "high",
    "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-5.3: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.