
GLM-5 / Chat
GLM-5 is Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. Built for expert developers, it delivers production-grade performance on large-scale programming tasks, rivaling leading closed-source models. With advanced agentic planning, deep backend reasoning, and iterative self-correction, GLM-5 moves beyond code generation to full-system construction and autonomous execution.
Press Enter to add, Backspace to remove
Playground Chat
Starten Sie eine Unterhaltung mit dem KI-Modell. Sie können alles fragen.
KI-generierte Antworten können in der Genauigkeit variieren.
GLM-5Toggleable Deep Thinking with Tools
GLM-5 is Zhipu’s thinking-capable chat model with thinking.type enable/disable, clear_thinking history cleanup, up to 128 function tools, and streaming chat completions on an OpenAI-compatible surface.

At a Glance
GLM-5 Thinking Stack
Deep reasoning when you need it — plain chat when you do not. One model, two cognitive gears.

Thinking Toggle
thinking.type switches between enabled and disabled so you can trade analytical depth for latency per request without swap models.

Clear Thinking History
thinking.clear_thinking (default true) cleans historical reasoning_content so multi-turn context stays compact and free of stale intermediate traces.

Function Tools
Register up to 128 function tools with tool_choice auto for agent loops, structured actions, and grounded retrieval handoffs.

Sampling & Stop
Temperature [0,1] default 1, top_p [0.01,1] default 0.95, max_tokens up to 131072, and a single stop word for clean parser boundaries.
How It Works
From request to reasoned answer in four moves.
Set Thinking
Enable or disable thinking.type based on task complexity and latency budget.
Attach Tools
Register function tools (up to 128) with tool_choice auto for agent loops.
Stream Tokens
stream defaults to true and returns text/event-stream for progressive paint.
Shape Output
Tune temperature (0–1, default 1), top_p (0.01–1, default 0.95), and one stop word.
GLM-5 Domains
Where thinking depth is a product dial, not a model swap.
Complex Analysis
Thinking on for multi-hop technical reasoning.
Agent Workflows
Function tools with tool_choice auto.
Fast Chat UX
Thinking off for lightweight conversational turns.
Structured Output
Temperature 0–1 and single stop word for parsers.
Prompt Tips
Get better first-pass results from GLM-5.
Leave thinking enabled for proofs and multi-document work; disable it for simple labels and rewrites.
GLM-5 currently accepts a single stop word — design parsers around that limit.
Ask for JSON or numbered steps when you will parse output programmatically.
GLM-5 Quickstart
Thinking chat with function tools.
curl -X POST "https://api.powertokens.ai/v1/chat/completions" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "glm-5",
"messages": [
{"role": "user", "content": "Outline a careful multi-step plan to migrate a REST API to versioned endpoints."}
],
"stream": true,
"temperature": 0.7
}'Technical Specifications
Confirmed parameters and runtime execution protocols.
Preisdetails
Die tatsächliche Abrechnung für dieses Modell wird dynamisch basierend auf den spezifischen Parametern Ihrer API-Anfrage berechnet. Nachfolgend finden Sie die spezifischen Kombinationen und ihre entsprechenden Preise:
| Modalität | Eingabe-Guthaben | Ausgabe-Guthaben | Eingabepreis | Ausgabepreis | Impliziter Cache-Treffer | Expliziter Cache-Treffer | Cache-Erstellung |
|---|---|---|---|---|---|---|---|
| Standard | 970/ 1M Tokens | 3,104/ 1M Tokens | $0.970 | $3.104 | $0.194 194/ 1M Tokens | $0.194 194/ 1M Tokens | -- |
Models from the Same Channel
Explore complementary models and alternative versions from the same provider channel.


GLM-5.3
glm-5.3
Zhipu AI's flagship text model optimized for complex software engineering and long-horizon Agent tasks. Features a 1M-token context window with mandatory reasoning (3 levels: low/high/max). Coding capability improved 50% over GLM-5.2 on Z.ai Code Bench; scores SOTA on Terminal Bench 3.0. Excels in cybersecurity tasks (cyber vulnerability discovery)


GLM-5.3 Flash
glm-5.3-flash
Z.ai's native multimodal model (320B-A18B) with text, image, video, and file input. Features hybrid sparse + linear attention, reducing attention compute 3x and KV cache 4.4x vs. GLM-5.3. Scores 57 on Artificial Analysis Intelligence Index—matching Claude Opus 4.8—with 1M context and 128K max output. Excels in agentic coding, vision-based Computer Use, and CAD reproduction


GLM-5.2
glm-5.2
Flagship text model purpose-built for long-horizon agentic workflows. Features a 1M context window supporting project-level engineering in a single session. Excels at autonomous coding: can complete development, testing, and multi-platform deployment from a single prompt. Top open-weight model per Artificial Analysis; #1 globally on Code Arena. MIT-licensed and Day-0 optimized for domestic AI chips


GLM-5 Turbo
glm-5-turbo
GLM-5 Turbo is a new model from Z.ai designed for fast inference and strong performance in agent-driven environments such as OpenClaw scenarios. It is deeply optimized for real-world agent workflows involving long execution chains, with improved complex instruction decomposition, tool use, scheduled and persistent execution, and overall stability across extended tasks.
Recommended Related Models
Explore complementary video and multimodal models with your unified API key.


MiniMax-M3.1-Flash
MiniMax-M3.1-Flash
Next-gen efficiency-focused multimodal model (428B MoE) for agentic workflows, coding, and computer use . Features a 1M-token context window and tunable thinking depth (low–max) . Scores 73.7% on DeepSWE v1.1, 61.7 on OSWorld V2, and 82% on CyberGym, while delivering 200+ TPS output speed at 70% lower agent-task cost than its predecessor


MiMo-V2.6-Pro-UltraSpeed
mimo-v2.6-pro-ultraspeed
High-speed edition of Xiaomi's 1T-parameter flagship, delivering roughly 10x output speed while matching the original model's quality. Features a 1M-token context window and native multimodal input (text, image, video, audio). Optimized for latency-sensitive agentic workflows, coding, and long-horizon tasks


Qwen3.8 Flash
qwen3.8-flash
Alibaba's cost-efficient multimodal reasoning model. Supports text, image, and video inputs with text output. Features a native 1M-token context window for long documents, codebases, and agentic workflows. Excels in coding assistance, desktop interaction, chart analysis, and long-video understanding. Compatible with OpenAI/Anthropic protocols for seamless integration


DeepSeek-V4.1-Flash
deepseek-v4-1-flash-260910
Lightweight flagship of DeepSeek's new Causal Encoder-Decoder family. 552B-parameter MoE with asymmetric activation—8B for input, 16B for output—delivering flagship-level intelligence at significantly lower inference cost . Native multimodal vision understanding. 1M-token context window with 384K max output. KV cache compressed to 1/4 HBM and 1/8 SSD of previous generation . Outperforms V4-Pro on Terminal-Bench 2.1 (90.6), DeepSWE v1.1 (74.2), and CyberGym (88.1)
Frequently Asked Questions
Everything you need to know before integrating this model.
Start Building with GLM-5 Today
Create an account in seconds to receive free test credits. No upfront payment or long-term contract required.