GLM 5.2
z-ai/glm-5.2- Z.ai
- Family · GLM 5
- Released 16 Jun 2026
- Open weights
- Text → Text
- Reasoning · adjustable
- Tools · JSON
Not on OSMEXO yet
Spec & list price
Context1M
Max output944K
Input / 1M$0.006
Output / 1M$6.801133.3× input
Scores · standing
Intelligence33.7#54/489
Coding68.8#29/196
Usage
Tokens / day97.5B▼ 6.5% 7d
Usage rank#32
Requests / day2.5M
Tool calls / req0.45
Per request · 30 days
Avg input24.4K1.6× median
Avg output727about median
Cache hit83%1.2× median
Reasoning share51%about median
Tool-call failures0.7%0.4× median
Consumption
Peak 358.8B 7 OctLow 64.4B 31 OctRank today #32
News & trends
11 mentions this week ▼ 31%this model · videos, posts, articles · 30 days
Trending · 14 days
- Запуск ИИ модели GLM 5.2 без видеокарты на ноутбуке238.3K views
- GLM 5.2 ranked below GLM 5. Where would you rank DeepSeek V4.1flash?10 upvotes
- Gemini 3 Preview was moved to unlisted due to a lack of reviews. Where would you rank GLM 5.2?8 upvotes
- GLM 5.2 launches with huge context and local run for coding tasks #GLM5 #launch #opensourceAI165 views
News
- vllm v0.29.0vLLM · 9 Sep
- transformers v5.16.1: Release v5.16.126 Aug
- sglang v0.5.16SGLang · 25 Jul
- sglang v0.5.15.post1SGLang · 14 Jul
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across 33 providers, latest day
Output speed#134/33378 tok/s1.2× faster than median
Fastest provider246 tok/sbest of 33 providers
First token#1621.01 sabout median
#4 cheapest of 9 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Family progress
Intelligence Index
- Family releases44.8
Technical details
- Reasoning
- on · high
- Tokenizer
- Other
- License
- MIT
- Parameters
- 753B
- Moderation
- none
- Downloads · likes
- 270.5K · 5.2K
Reasoning effort
- xhigh
- high · default
- none
Request parameters 21
- tools
- tool_choice
- parallel_tool_calls
- response_format
- structured_outputs
- reasoning
- reasoning_effort
- include_reasoning
- temperature
- top_p
+11 more
- top_k
- min_p
- max_tokens
- stop
- seed
- frequency_penalty
- logit_bias
- logprobs
- presence_penalty
- repetition_penalty
- top_logprobs
When to use
Best for
- Very long document and codebase work
- Agentic coding with tool calling
- Large outputs such as full reports
- Reasoning toggled on or off per request
Think twice for
- Image input
- Fast cheap chat
- Browser-driving agents
Research papers
- Hierarchical Skill Evolution for Coding Agent26 Oct
- RepoTrace: Improving Requirement Tracing in Multi-Round Agent Interactions26 Oct
- Building Presentation Intelligence for High-Quality Slide Generation22 Oct
- Joules-to-GDP: How Much Economic Value Do AI Systems Produce per Unit of Energy?22 Oct
- VeriEvolve: Evaluator-First Evolution for Tool-Using Agents21 Oct
- HarnessPro: Recursive Harness Self-Improvement with Pareto Optimization for Performance-Cost Trade-offs19 Oct
- Peer Pressure: Execution-Verified Co-Evolution of Attacker and Defender Populations for Code Security17 Oct
- Where Natural Language Ends in Agents: Executable Skill Composition with MetaSkill17 Oct
Social
Most watched and discussed · 90 days
- Запуск ИИ модели GLM 5.2 без видеокарты на ноутбукеВлад Нейронов🤖 · 238.3K views · 26 Oct
- GLM 5.2 launches with huge context and local run for coding tasks #GLM5 #launch #opensourceAIEli Sadler · 165 views · 22 Oct
- An Italian developer ran a 744 billion parameter GLM 5.2 model locally without a GPU using ColibriPhineas and Ferb Tech · 92 views · 23 Oct
- GLM 5.2: 300 Million Free Tokens Daily? 🤯 The AI Coding Revolution!codecbyte · 53 views · 24 Oct
- Self-Hosting GLM-5.2 with llama.cppSyntaxMedia · 48 views · 26 Oct
- GLM-5.2 сама дісталась усіх 8 пристроївNotebookLM Stories · 32 views · 22 Oct
- Reflection AI Beam: The Open Model That Beat GLM-5.2AGENT AMIGOS · 22 views · 29 Oct
- Stop Comparing GLM 5.2 to Video ToolsStackLab · 18 views · 20 Oct
- Zhipu AI brengt GLM-5.2 uit, open model met 744 miljard parametersaiMattie · 8 views · 25 Oct
- #54 SpaceX Acquire Cursor, Framer 3.0, v0 Design Mode, GLM-5.2, SnapCmd+AI · 8 views · 20 Oct
- GLM 5.2, WordPress, Nestlé y benchmark: 7 lecciones de negocio e IAMomentazos · 7 views · 27 Oct
- Китайская GLM 5.2 против Claude Codedomidva · 7 views · 24 Oct
Same family
GLM 5- 44.8
- 26.1
- 27.9
Alternatives
other vendors · similar capability- C33.9
- A35.2
- 34.8
- I30.5
Want GLM 5.2 on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.