Z.ai · Open weights
GLM 5.2 Thinking TEE
GLM-5.2 with thinking enabled for harder long-horizon coding, autonomous agent workflows, and complex engineering tasks. Running inside a TEE (Trusted Execution Environment), with attestation support.
- Reasoning
- Tools
- Structured output
- Input
- Text
- Output
- Text
- Reasoning effort
- low · medium · high · xhigh · max
- Discovery tier
- Open weights
Pricing
USD per 1M tokens- Input
- $1.40
- Output
- $4.40
- Cache read
- $0.70
Reference prices as published by the model provider. API example
Specifications
- Context window
- 1M
- Max output
- 131.1K
- Average output (7 days)
- 2K tokens
- Longer than 92% of 468 models
Side by side
GLM 5.2 Thinking TEE compared
Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.
| GLM 5.2 Thinking TEEZ.aiThis model | DeepSeek V4 Flash Vision ExpDeepSeekRemove | |
|---|---|---|
| Provider | Z.ai | DeepSeek |
| Discovery tier | Open weights | Efficient |
| Added | Jun 16, 2026 | Aug 21, 2026 |
| Limits | ||
| Context window | 1.05M | 1.05M |
| Max output | 131.07K | 384KBest |
| Average output | 2K tokens | 1.3K tokens |
| Pricing, USD per 1M tokens | ||
| Input | $1.40 | $0.22Best |
| Output | $4.40 | $0.65Best |
| Cache read | $0.70 | $0.0069Best |
| Performance | ||
| Throughput | — | 155.2 tok/s |
| Time to first token | — | 2.1 s |
| Request success (24h) | 87.5% | 99.7%Best |
| Benchmark indexes | ||
| Intelligence Index | 33.7 | 34.3Best |
| Coding Index | 68.8 | 69.1Best |
| Capabilities | ||
| Reasoning | Yes | Yes |
| Tools | Yes | Yes |
| Vision | No | Yes |
| Structured output | Yes | Yes |
| Reasoning effort | low, medium, high, xhigh, max | none, low, high, max |
Public evaluations
Benchmarks
Independent evaluation results, with how each score ranks among the models in the Noumen catalog.
- GLM 5.2 Thinking TEE
- DeepSeek V4 Flash Vision Exp
Intelligence Index
- GLM 5.2 Thinking TEE33.7
- DeepSeek V4 Flash Vision Exp34.3
GLM 5.2 Thinking TEE tested at max reasoning effort
GLM 5.2 Thinking TEE: Better than 76% of 326 scored models
Indexes
Coding Index
Composite of the coding evaluations
- GLM 5.2 Thinking TEE: 68.8
- DeepSeek V4 Flash Vision Exp: 69.1Top
Agentic work
AutomationBench-AA
Workflow automation, penalized for guardrail violations
- GLM 5.2 Thinking TEE: 28.4%
- DeepSeek V4 Flash Vision Exp: 54.0%Top
AutomationBench-AA tasks completed
Workflows finished end to end without a guardrail violation
- GLM 5.2 Thinking TEE: 6.2%
- DeepSeek V4 Flash Vision Exp: 23.0%Top
AA-Briefcasev1.1
Agentic knowledge work, rated head to head
- GLM 5.2 Thinking TEE: 1,227 Elo
- DeepSeek V4 Flash Vision Exp: 1,257 EloTop
GDPval-AAv2.1
Economically valuable occupational tasks, rated head to head
- GLM 5.2 Thinking TEE: 1,373 Elo
- DeepSeek V4 Flash Vision Exp: 1,433 EloTop
Document reasoning
GDP.pdf
Professional reasoning over PDF documents, all-pass rate
- GLM 5.2 Thinking TEE: 10.4%
- DeepSeek V4 Flash Vision Exp: 11.0%Top
AA-LCRv1.1
Reasoning across long documents
- GLM 5.2 Thinking TEE: 78.3%
- DeepSeek V4 Flash Vision Exp: 79.7%Top
MLCR-AA
Long-context reasoning over medical records
- GLM 5.2 Thinking TEE: 12.8%
- DeepSeek V4 Flash Vision Exp: —
Reasoning and knowledge
Humanity’s Last Exam
Expert-level questions across academic fields
- GLM 5.2 Thinking TEE: 41.1%Top
- DeepSeek V4 Flash Vision Exp: 38.6%
GPQA Diamond
Graduate-level science questions
- GLM 5.2 Thinking TEE: 89.5%
- DeepSeek V4 Flash Vision Exp: 90.8%Top
IFBench
Following precise output instructions
- GLM 5.2 Thinking TEE: 73.3%
- DeepSeek V4 Flash Vision Exp: —
Coding
Terminal-Benchv4.0
Practical software engineering in a terminal
- GLM 5.2 Thinking TEE: 1.0%
- DeepSeek V4 Flash Vision Exp: 12.1%Top
SciCode
Scientific computing in Python
- GLM 5.2 Thinking TEE: 51.2%Top
- DeepSeek V4 Flash Vision Exp: 50.3%
Legacy evaluations
Terminal-Bench Hard
Earlier terminal task suite
- GLM 5.2 Thinking TEE: 50.8%
- DeepSeek V4 Flash Vision Exp: —
τ²-Bench Telecom
Tool use in a customer-support simulation
- GLM 5.2 Thinking TEE: 99.1%
- DeepSeek V4 Flash Vision Exp: —
AA-LCR (unversioned)
Long-context reasoning, earlier grading
- GLM 5.2 Thinking TEE: 78.3%
- DeepSeek V4 Flash Vision Exp: 79.7%Top
Scores from Artificial Analysis, updated Oct 26, 2026. GLM 5.2 Thinking TEE 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.
- Prompt cache hits
- 45.3%
- About $1.08 per 1M input tokens with caching
- Request success (24h)
- 87.5%
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.
Developers
Use it through the API
Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.
curl https://www.noumen.xyz/api/v1/chat/completions \
-H "Authorization: Bearer $NOUMEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "TEE/glm-5.2: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="TEE/glm-5.2:thinking",
messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)Z.ai
Related models
Other Z.ai models, starting with the same family.