LIVE
Study measures how coding agents pick third-party tools03/09/26|OpenAI's GPT-6 Astra sets new state-of-the-art on ARC-AGI-303/09/26 · OpenAI|ESPO: A Prompt Optimization Method That Beats GEPA With Shorter, More Stable Prompts03/09/26|"Last Translation Benchmark": a large-scale collaborative effort for a definitive MT benchmark03/09/26|NVIDIA pushes local AI at IFA 2026 with RTX Spark PCs and a personal inference router03/09/26 · NVIDIA|Google DeepMind unveils WeatherNext 3, its most accurate weather AI model yet03/09/26 · Google DeepMind|OpenAI Launches $1B Daybreak Program for Frontline Cybersecurity Defenders03/09/26 · OpenAI|Hcompany releases NeoMME, a compact multimodal-native encoder for retrieval03/09/26 · Hcompany|Playco cuts manual fixes by 50% using GPT-6 Astra for prototyping03/09/26 · OpenAI|Legora reviews 41 financial documents in minutes with GPT-6 Astra03/09/26 · OpenAI|Hugging Face open-sources a project training a coding model to paint watercolours03/09/26 · Hugging Face|A 350M-Parameter Model Improved in 100 GRPO Steps for Better Structured Outputs03/09/26|Study measures how coding agents pick third-party tools03/09/26|OpenAI's GPT-6 Astra sets new state-of-the-art on ARC-AGI-303/09/26 · OpenAI|ESPO: A Prompt Optimization Method That Beats GEPA With Shorter, More Stable Prompts03/09/26|"Last Translation Benchmark": a large-scale collaborative effort for a definitive MT benchmark03/09/26|NVIDIA pushes local AI at IFA 2026 with RTX Spark PCs and a personal inference router03/09/26 · NVIDIA|Google DeepMind unveils WeatherNext 3, its most accurate weather AI model yet03/09/26 · Google DeepMind|OpenAI Launches $1B Daybreak Program for Frontline Cybersecurity Defenders03/09/26 · OpenAI|Hcompany releases NeoMME, a compact multimodal-native encoder for retrieval03/09/26 · Hcompany|Playco cuts manual fixes by 50% using GPT-6 Astra for prototyping03/09/26 · OpenAI|Legora reviews 41 financial documents in minutes with GPT-6 Astra03/09/26 · OpenAI|Hugging Face open-sources a project training a coding model to paint watercolours03/09/26 · Hugging Face|A 350M-Parameter Model Improved in 100 GRPO Steps for Better Structured Outputs03/09/26|
🇨🇳 China💻 CodeReleased August 2026

GLM-5.3

Zhipu AI
Open-sourceSelf-host
Verdict

A solid pick for self-hosting and full data control.

Overview

GLM-5.3 shares its base with GLM-5.2 but redirects post-training toward complex coding and long-horizon agentic work, gaining 50% on Z.ai's internal Code Bench and posting an open-weights SOTA on Terminal Bench 3.0. Its emergent cyber capability is strong for vulnerability discovery, though that same strength invites caution over dual use. Limit: on several agentic benchmarks it still trails GPT-5.6 Sol. Open weights under a commercial licence allow on-premise EU deployment with data staying local, but the 753-billion-parameter size demands serious server-grade GPU capacity.

Skill profile

Not disclosed

Strengths

  • Very long context
  • Open-source and self-hostable

Limitations

  • No native GDPR guarantee
  • API pricing not disclosed

Who is it for

A good fit if…
  • you build with a coding agent
  • you want to control cost or self-host
Skip it if…
  • you handle sensitive EU data

Ideal use cases

  • Complex coding agents
  • DevSecOps automation
  • Vulnerability discovery
  • Long-horizon multi-step tasks
  • Sovereign on-prem deployment

Access & availability

Self-hostable (open weights)

Key specifications

Context
1M
Input price
Output price
Speed
Price not auditedBenchmarks not audited

Privacy

Not GDPR-compliant

Run it locally

Dedicated GPU server304k2k
QuantizationDiskRAM / VRAMTypical hardware
Q4 · recommended429.2 GB474 GBServer GPU infra (H100/A100…)
Q8 · balanced805.7 GB888 GBServer GPU infra (H100/A100…)
FP16 · max quality1506 GB1659 GBServer GPU infra (H100/A100…)

Estimates for a moderate context. Long contexts need more RAM (KV cache).

Deploy

Copy-ready commands generated from this card. Adjust context length and GPU count to your hardware.

Estimated memory · FP16 1659 Go · Q4 474 Go2 GPUs

OpenAI-compatible server for production on NVIDIA GPUs.

pip install vllm
vllm serve zai-org/GLM-5.3 \
  --max-model-len 32768 \
  --tensor-parallel-size 2 \
  --dtype auto
zai-org/GLM-5.3
Advanced data · for experts
Architecture, modalities, detailed cost, full benchmarks
Architecture
MoE
Size
753B (MoE, 40B actifs)
Cutoff date
Inputs
text
Outputs
text
License
glm-5.3
Hosting
Hallucination score
Estimated cost (API)
Typical exchange (~3k in / 1k out)
Not disclosed
1M in + 1M out
Not disclosed
All benchmarks
Arena Elo
MMLU
GPQA
HumanEval
SWE-Bench
MATH

Similar models