Back to Developer ZoneOfficial Pricing Official Pricing Official Pricing Official Pricing Official Pricing Official Pricing
Zhipu AI Models
Explore all 6 models from Zhipu AI with detailed pricing, pros & cons, and developer recommendations.
6
Models
$0.050
Lowest Input
1M
Max Context
3
Quality Tiers
Quick Recommendations
Best Value: GLM-4-Flash ($0.050/1M)
Best Quality: GLM-5.1
GLM-5.1
FlagshipComplex coding, long-horizon agentic tasks, open-source deployment
When to use: Open-source coding assistant, internal developer tooling, agentic coding workflows, and teams needing self-hosted frontier-capable models.
Upgrade Highlights
- ◆754B MoE open-weight — MIT license, full commercial use
- ◆SWE-bench matches GPT-5.4 — frontier coding performance
- ◆8-hour autonomous task execution on a single problem
- ◆Rumination: iterative internal reasoning for correctness
- ◆Self-host on your own GPUs — no vendor lock-in
Input Price
$0.830
per 1M tokens
Output Price
$3.31
per 1M tokens
Cached Input
$0.170
per 1M tokens
Batch Input
—
per 1M tokens
Context Window: 1M
Max Output: 16,384 tokens
Knowledge Cutoff: 2026-04
VisionFunction CallingFine-tuningJSON ModeFree Tier
Pros
- 754B MoE open-weight (MIT license)
- Matches GPT-5.4 on SWE-bench coding
- 8-hour sustained autonomous task execution
- Self-hostable with full commercial rights
- Rumination architecture for deep reasoning
Cons
- 754B params requires substantial GPU infrastructure to self-host
- Weaker English vs closed frontier models on generalist tasks
- No vision on base model
Performance
Output Speed~40 tok/s
Rate Limit3,000 RPM
Multimodal
Image InputImage OutputAudio InputAudio Output
Benchmarks
MMLU (CN)
91.2%
C-Eval
93.5%
SWE-Verified
78.6%
CMMLU
92.1%
GLM-4.6
FlagshipChinese language tasks, enterprise AI
When to use: Chinese-language enterprise applications, customer service bots, and content generation targeting Chinese markets.
Upgrade Highlights
- ◆Top-tier Chinese NLU and generation — beats GPT-4 on Chinese benchmarks
- ◆128K context with 16K max output — longest output in class
- ◆Full function calling for agent workflows
- ◆Fine-tuning available for domain adaptation
- ◆$0.50/$2.00 — competitive with GPT-4o at half the price
Input Price
$0.500
per 1M tokens
Output Price
$2.00
per 1M tokens
Cached Input
$0.100
per 1M tokens
Batch Input
—
per 1M tokens
Context Window: 128K
Max Output: 16,000 tokens
Knowledge Cutoff: 2025-03
VisionFunction CallingFine-tuningJSON ModeFree Tier
Pros
- Best Chinese language performance
- 128K context, 16K output
- Strong function calling
- Fine-tuning support
Cons
- Weaker English vs GPT-4
- No vision on base model
- Smaller ecosystem
Performance
Output Speed~60 tok/s
Rate Limit5,000 RPM
Multimodal
Image InputImage OutputAudio InputAudio Output
Benchmarks
MMLU (CN)
84.5%
C-Eval
89.2%
CMMLU
88.7%
GLM-4.5
Mid-tierBalanced Chinese/English tasks
When to use: Bilingual applications needing good Chinese and English at mid-tier pricing.
Upgrade Highlights
- ◆Strong bilingual: competitive in both Chinese and English
- ◆128K context at $0.30/1M — affordable long-context
- ◆16K max output for long-form generation
- ◆Fine-tuning support for customization
Input Price
$0.300
per 1M tokens
Output Price
$1.20
per 1M tokens
Cached Input
$0.080
per 1M tokens
Batch Input
—
per 1M tokens
Context Window: 128K
Max Output: 16,000 tokens
Knowledge Cutoff: 2025-03
VisionFunction CallingFine-tuningJSON ModeFree Tier
Pros
- Strong bilingual performance
- 128K context
- 16K max output
- Cost-effective
Cons
- Less capable than GLM-4.6
- No vision
- Smaller model ecosystem
Performance
Output Speed~75 tok/s
Rate Limit8,000 RPM
Multimodal
Image InputImage OutputAudio InputAudio Output
Benchmarks
MMLU
76.8%
C-Eval
83.1%
GLM-4-Plus
Mid-tierGeneral purpose, API integration
When to use: General-purpose API integration, chatbots, and content generation at budget-friendly pricing.
Upgrade Highlights
- ◆Versatile mid-tier model for most use cases
- ◆128K context at just $0.20/1M input
- ◆Full function calling for tool use
- ◆Fine-tuning available
Input Price
$0.200
per 1M tokens
Output Price
$0.800
per 1M tokens
Cached Input
$0.050
per 1M tokens
Batch Input
—
per 1M tokens
Context Window: 128K
Max Output: 8,192 tokens
Knowledge Cutoff: 2025-03
VisionFunction CallingFine-tuningJSON ModeFree Tier
Pros
- Good all-rounder
- 128K context
- Affordable pricing
- Function calling
Cons
- 8K max output
- No vision
- Weaker on complex reasoning
Performance
Output Speed~85 tok/s
Rate Limit10,000 RPM
Multimodal
Image InputImage OutputAudio InputAudio Output
Benchmarks
MMLU
73.5%
C-Eval
79.8%
GLM-4-Flash
LiteHigh-throughput, low-latency tasks
When to use: High-volume tasks like classification, summarization, and simple Q&A where speed and cost matter.
Upgrade Highlights
- ◆Fastest GLM model — optimized for throughput
- ◆$0.05/1M input — ultra-budget friendly
- ◆128K context despite lite tier
- ◆Free tier: 1M tokens/day for development
Input Price
$0.050
per 1M tokens
Output Price
$0.200
per 1M tokens
Cached Input
$0.010
per 1M tokens
Batch Input
—
per 1M tokens
Context Window: 128K
Max Output: 8,192 tokens
Knowledge Cutoff: 2025-03
VisionFunction CallingFine-tuningJSON ModeFree Tier
Pros
- Extremely fast inference
- 128K context
- Very low cost
- Free tier available
Cons
- Basic reasoning only
- No fine-tuning
- No vision
Performance
Output Speed~200 tok/s
Rate Limit30,000 RPM
Multimodal
Image InputImage OutputAudio InputAudio Output
Benchmarks
MMLU
65.2%
C-Eval
72.1%
GLM-4V-Plus
Mid-tierChinese multimodal, document AI
When to use: Chinese document analysis, receipt/invoice processing, and visual Q&A for Chinese markets.
Upgrade Highlights
- ◆Native multimodal with strong Chinese OCR
- ◆Document AI: receipts, invoices, forms
- ◆Visual Q&A optimized for Chinese content
- ◆Function calling for multimodal agent workflows
Input Price
$0.300
per 1M tokens
Output Price
$1.20
per 1M tokens
Cached Input
$0.080
per 1M tokens
Batch Input
—
per 1M tokens
Context Window: 8K
Max Output: 4,096 tokens
Knowledge Cutoff: 2025-03
VisionFunction CallingFine-tuningJSON ModeFree Tier
Pros
- Native vision-language
- Strong Chinese OCR
- Document and chart understanding
- Function calling
Cons
- 8K context only
- 4K max output
- No fine-tuning
Performance
Output Speed~50 tok/s
Rate Limit3,000 RPM
Multimodal
Image InputImage OutputAudio InputAudio Output
Benchmarks
MMMU (CN)
62.8%
DocVQA
85.3%
Side-by-Side Comparison
| Model | Tier | Input | Output | Context |
|---|---|---|---|---|
| GLM-5.1 | Flagship | $0.830 | $3.31 | 1M |
| GLM-4.6 | Flagship | $0.500 | $2.00 | 128K |
| GLM-4.5 | Mid-tier | $0.300 | $1.20 | 128K |
| GLM-4-Plus | Mid-tier | $0.200 | $0.800 | 128K |
| GLM-4-Flash | Lite | $0.050 | $0.200 | 128K |
| GLM-4V-Plus | Mid-tier | $0.300 | $1.20 | 8K |