Z.ai · Open weights
GLM 5.3 TEE
GLM-5.3 is Z.AI's open-weight reasoning model for complex software engineering, autonomous agents, vulnerability research, and long-horizon tasks. Reasoning is always on; choose low, high, or max reasoning effort (defaults to max). This text-only TEE deployment is verified through the selected provider: Redpill attestation with signed completion receipts or the official Tinfoil SDK's ATC/EHBP verification.
- Reasoning
- Tools
- Structured output
- Parallel tools
- Input
- Text
- Output
- Text
- Reasoning effort
- low · high · max
- Discovery tier
- Open weights
Pricing
USD per 1M tokens- Input
- $1.40
- Output
- $4.40
- Cache read
- $0.26
Reference prices as published by the model provider. API example
Specifications
- Context window
- 1M
- Max output
- 131.1K
- Average output (7 days)
- 1.2K tokens
- Longer than 77% of 465 models
Side by side
GLM 5.3 TEE compared
Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.
| GLM 5.3 TEEZ.aiThis model | GLM 5.2 ThinkingZ.aiRemove | |
|---|---|---|
| Provider | Z.ai | Z.ai |
| Discovery tier | Open weights | Open weights |
| Added | Aug 31, 2026 | Jun 15, 2026 |
| Limits | ||
| Context window | 1.05M | 1.05M |
| Max output | 131.07K | 131.07K |
| Average output | 1.2K tokens | 1.8K tokens |
| Pricing, USD per 1M tokens | ||
| Input | $1.40 | $0.42Best |
| Output | $4.40 | $1.32Best |
| Cache read | $0.26 | $0.078Best |
| Performance | ||
| Throughput | 48.0 tok/s | 108.8 tok/sBest |
| Time to first token | 4.6 s | 1.8 sBest |
| Request success (24h) | 99.7% | 99.7% |
| Benchmark indexes | ||
| Intelligence Index | 44.8Best | 33.7 |
| Coding Index | 74.8Best | 68.8 |
| Capabilities | ||
| Reasoning | Yes | Yes |
| Tools | Yes | Yes |
| Structured output | Yes | Yes |
| Parallel tools | Yes | No |
| Reasoning effort | low, high, max | high, max |
Public evaluations
Benchmarks
Independent evaluation results, with how each score ranks among the models in the Noumen catalog.
- GLM 5.3 TEE
- GLM 5.2 Thinking
Intelligence Index
- GLM 5.3 TEE44.8
- GLM 5.2 Thinking33.7
GLM 5.3 TEE tested at max reasoning effort
GLM 5.3 TEE: Better than 90% of 326 scored models
Indexes
Coding Index
Composite of the coding evaluations
- GLM 5.3 TEE: 74.8Top
- GLM 5.2 Thinking: 68.8
Agentic work
AutomationBench-AA
Workflow automation, penalized for guardrail violations
- GLM 5.3 TEE: 62.2%Top
- GLM 5.2 Thinking: 28.4%
AutomationBench-AA tasks completed
Workflows finished end to end without a guardrail violation
- GLM 5.3 TEE: 29.4%Top
- GLM 5.2 Thinking: 6.2%
Harvey LAB-AAv1.1
Legal tasks passed in full, with no material hallucination
- GLM 5.3 TEE: 0.3%
- GLM 5.2 Thinking: —
AA-Briefcasev1.1
Agentic knowledge work, rated head to head
- GLM 5.3 TEE: 1,505 EloTop
- GLM 5.2 Thinking: 1,227 Elo
GDPval-AAv2.1
Economically valuable occupational tasks, rated head to head
- GLM 5.3 TEE: 1,649 EloTop
- GLM 5.2 Thinking: 1,373 Elo
Document reasoning
GDP.pdf
Professional reasoning over PDF documents, all-pass rate
- GLM 5.3 TEE: 11.2%Top
- GLM 5.2 Thinking: 10.4%
AA-LCRv1.1
Reasoning across long documents
- GLM 5.3 TEE: 79.7%Top
- GLM 5.2 Thinking: 78.3%
MLCR-AA
Long-context reasoning over medical records
- GLM 5.3 TEE: 17.2%Top
- GLM 5.2 Thinking: 12.8%
Reasoning and knowledge
Humanity’s Last Exam
Expert-level questions across academic fields
- GLM 5.3 TEE: 42.3%Top
- GLM 5.2 Thinking: 41.1%
GPQA Diamond
Graduate-level science questions
- GLM 5.3 TEE: 91.7%Top
- GLM 5.2 Thinking: 89.5%
IFBench
Following precise output instructions
- GLM 5.3 TEE: —
- GLM 5.2 Thinking: 73.3%
Coding
Terminal-Benchv4.0
Practical software engineering in a terminal
- GLM 5.3 TEE: 41.9%Top
- GLM 5.2 Thinking: 1.0%
SciCode
Scientific computing in Python
- GLM 5.3 TEE: 59.0%Top
- GLM 5.2 Thinking: 51.2%
Legacy evaluations
Terminal-Bench Hard
Earlier terminal task suite
- GLM 5.3 TEE: —
- GLM 5.2 Thinking: 50.8%
τ²-Bench Telecom
Tool use in a customer-support simulation
- GLM 5.3 TEE: —
- GLM 5.2 Thinking: 99.1%
AA-LCR (unversioned)
Long-context reasoning, earlier grading
- GLM 5.3 TEE: 79.7%Top
- GLM 5.2 Thinking: 78.3%
Scores from Artificial Analysis, updated Oct 26, 2026. GLM 5.3 TEE was tested at max reasoning effort; compare scores measured at the same effort.Artificial Analysis
- GLM 5.3 TEE
- GLM 5.2 Thinking
| Leaderboard | GLM 5.3 TEE | GLM 5.2 Thinking |
|---|---|---|
| Industry Legal And Government | Not in top 20 | #18 · 1491.1 |
| Spanish | Not in top 20 | #16 · 1478.5 |
Ranks among Noumen catalog models on LMArena leaderboards. Only top-20 placements are listed.LMArena
- GLM 5.3 TEE
- GLM 5.2 Thinking
| Leaderboard | GLM 5.3 TEE | GLM 5.2 Thinking |
|---|---|---|
| FrontierMath v2 (Tiers 1–3) | Not in top 20 | #7 · 59.2% |
| SWE-bench Verified | Not in top 20 | #3 · 78.7% |
| GPQA Diamond | Not in top 20 | #6 · 91.9% |
Ranks among Noumen catalog models on Epoch AI leaderboards. Only top-20 placements are listed.Epoch AI
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
- 48.0 tok/s
- Faster than 35% of 250 models
- Time to first token
- 4.6 s
- Quicker than 40% of 250 models
- Prompt cache hits
- 87.5%
- About $0.40 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.
- 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.
curl https://www.noumen.xyz/api/v1/chat/completions \
-H "Authorization: Bearer $NOUMEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "TEE/glm-5.3",
"reasoning_effort": "low",
"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-5.3",
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.