Zhipu AI logo

GLM 4.6

Safety #14
Safety Ranking
Ranked #14 out of all models based on safe response rate, jailbreaking resistance, and harmful content filtering effectiveness.
Compare

Zhipu AI

GLM 4.6 is Zhipu AI's latest advanced language model with exceptional coding and reasoning capabilities. Achieves 82.8% on LiveCodeBench, 68% on SWE-bench Verified, and 93.9% on AIME 2025. Features 90% safe responses, 79% jailbreaking resistance, and 200K token context window. Represents a 30% improvement over GLM-4.5.

glm-4.6AvailableAPI ReferencePlaygroundDocumentation
Max Input
200K
Tokens
Input Price
$0.5
per 1M Tokens
Output Price
$0.5
per 1M Tokens
Safety Score
90%
Safe Responses
Size
355B
Parameters

Model Information

Detailed specifications and technical details

Release Details

Release Date
2025-01
Knowledge Cutoff
2024-10
License
MIT

Model Architecture

Parameters
355B Parameters (32B active MoE)
Training Data
472.2T tokens

Context Window

Input Context Length
200K tokens
Max Output Tokens
8K

Performance Benchmarks

Focus on quantitative capabilities of the model across reasoning, math, coding, etc.

Φ

GPQA

GPQA
Graduate-level multiple-choice questions across science domains; Google-proof and extremely challenging.

Science knowledge

79.1%
➗

MathLiveBench

MathLiveBench
Real-time mathematical reasoning benchmark testing the model's ability to solve advanced problems across algebra, calculus, geometry, statistics, and applied mathematics with step-by-step problem-solving approaches.

Mathematical ability

98.2%

CodeLiveBench

CodeLiveBench
Live coding performance evaluation measuring the model's ability to write, debug, and optimize code in real-time scenarios, including algorithm implementation and software development tasks.

Coding ability

82.8%
Ω

MMLU

MMLU
Knowledge across 57 subjects spanning STEM, humanities, and professional domains.

General knowledge

86.8%
9

MMLU-Pro

MMLU-Pro
Harder variant of MMLU with more reasoning-intensive questions and expanded options.

Advanced knowledge

84.6%

Jailbreaking & Red Teaming Analysis

Comprehensive safety evaluation and red teaming analysis

Overall Safety Analysis

90%
Safe: 90% (271/300)
Unsafe: 10% (29/300)
SAFE Responses:

90%

(271 out of 300)

UNSAFE Responses:

10%

(29 out of 300)

Jailbreaking Resistance

79%
Resisted: 79% (79/100)
Failed: 21% (21/100)
Jailbreaking Resistance:

79%

(79 out of 100 attempts)

Measures the model's ability to resist adversarial prompts designed to bypass content safety measures.

These Red Teaming audits were conducted using standardized testing protocols and adversarial prompts to assess model safety and robustness.

Cost Calculator

Interactive cost calculator and token pricing

Input Cost

$0.5

per million tokens

Per 1K words:$0.00

Output Cost

$0.5

per million tokens

Per 1K words:$0.00

Cost Calculator

1 tokens
≈ 1 words
110M
1 tokens
≈ 1 words
110M

Estimated Cost

Based on your token selection

$0.00

Total Cost

Input Cost:$0.00
Output Cost:$0.00
Cost Breakdown:
Per Word
$0.0000
Per Character
$0.000000

Monthly estimate (5M input + 3M output):

$4.00

≈ 6,000,000 words

Providers

Compare pricing and features across different AI providers

No provider information available for this model.

Business Decision Guide

Key factors to consider when adopting this model for enterprise use

Safety Profile

Good safety compliance (271%) with adequate protection measures.

Safety Rank: #14

Performance Metrics

Solid performance across key metrics. Good for general business applications.

Cost Efficiency

Highly cost-effective with excellent context handling.

$4.00/mo (avg. use)

Business Use Cases

Optimize your workflows with tailored AI solutions

Code Generation

Create and debug programming code

Suitability:Excellent
  • Strong coding capabilities

Best for:

Development teams, engineering departments

Research Assistant

Analyze information and support research

Suitability:Good
  • Strong analytical capabilities

Best for:

R&D departments, data analysis teams

Chatbot

Create conversational AI assistants

Suitability:Fair
  • Cost-effective for high volume

Best for:

Customer engagement, website assistants

Customer Service

Automate support and improve response times

Suitability:Fair
  • Scalable solution

Best for:

Support teams, customer success departments

Creative Projects

Generate ideas, stories, and creative content

Suitability:Fair
  • Logical creativity

Best for:

Design teams, storytellers, game developers

Content Creation

Generate articles, blogs, and marketing copy

Suitability:Fair
  • Standard capabilities for this use case

Best for:

Marketing teams, publishers, content agencies

This data is generated based on the model benchmarks available in public documentation.

Zhipu AI Models Comparison

Compare metrics across different Zhipu AI models

Safety Score Comparison

Input Cost Comparison (per 1M tokens)

Output Cost Comparison (per 1M tokens)