How to Use GLM-5.2 in Codex
GLM-5.2 suits codebase audits, cross-file refactoring, and engineering tasks that need sustained context. Compared with GLM-5.1, it expands the context window from 200K to 1M tokens and adds control over reasoning effort.
Prompts to try with GLM-5.2 in Codex
Copy an example into Codex, or ask in your own words.
Use GLM-5.2 to audit this repository. Map modules, API contracts, and data flows; cite file paths for each finding.
Use GLM-5.2 to separate this repository's billing logic from request handling. Preserve public behavior, plan the changes, then run relevant tests.
Use GLM-5.2 to investigate the failing integration test. Reproduce it, test competing causes, and verify the smallest justified fix.
Use GLM-5.2 to migrate this project's deprecated API calls. Inventory callers, preserve compatibility, and report validation results.
Use GLM-5.2 to investigate this project's slowest endpoint. Establish a baseline, change the measured bottleneck, and rerun the benchmark.
How to set up GLM-5.2 in Codex
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the hosted server
codex mcp add spiral --url https://api.spiral.ai/mcp
- Sign in
Codex opens the browser: sign in to Spiral and click Authorize. No API key. codex mcp list then shows Spiral.
codex mcp login spiral
Use the Spiral skill
- Send Codex this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run GLM-5.2
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openrouter", "model": "z-ai/glm-5.2", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Codex finds GLM-5.2 on Spiral, shows the price and runs it.
GLM-5.2 providers and prices
Spiral runs GLM-5.2 through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.012916 per run at the default inputs. $0.806 per 1M input tokens, $3.146 per 1M output tokens.
Settings you can ask for
Ask Codex for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Min p | 0 to 1 | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 943718 | The most output tokens to generate. | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Reasoning effort | none, high, or xhigh | high | How much the model reasons. |
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 | ||
| Parallel tool calls | on or off |
Every API field of Z.ai: GLM 5.2 is on its tool page: open its card above.
What can you do with GLM-5.2 in Codex?
Turn project context into coordinated code changes with GLM-5.2. Give the agent a bounded goal, repository conventions, and explicit verification requirements.
Map an unfamiliar codebase
Ask GLM-5.2 to identify module responsibilities, API contracts, dependencies, and the constraints a future change must preserve.
Refactor across files
Give GLM-5.2 a bounded refactoring goal, protected behavior, and acceptance checks. Request a plan followed by implementation and verification.
Plan an API migration
Have GLM-5.2 compare existing and proposed contracts, order dependent changes, and define compatibility tests.
Investigate a complex bug
Supply GLM-5.2 with textual logs, reproduction steps, and relevant code. Request competing explanations and tests that distinguish them.
Optimize measured bottlenecks
Give GLM-5.2 profiling results and performance targets. Ask for a focused change and a repeatable before-and-after measurement.
Implement a detailed specification
Ask GLM-5.2 to turn requirements into an implementation plan, working code, and acceptance checks while tracking unresolved decisions.
More on GLM-5.2 and Codex
Overview
GLM-5.2 in other agents
More Codex guides
Questions about GLM-5.2 in Codex
Can Codex use GLM-5.2?
Yes. Add Spiral to Codex as an MCP server, and Codex can find and run GLM-5.2 like any built-in tool.
How much does GLM-5.2 cost in Codex?
Up to $0.036211 per run at the default inputs. $0.0091 per 1M input tokens, $8.84 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Codex plan is separate.
Do I need an OpenRouter account to use GLM-5.2 in Codex?
No. Spiral runs GLM-5.2 on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up GLM-5.2 in Codex?
About two minutes: add the hosted server, then sign in.
Does Codex wait while GLM-5.2 runs?
GLM-5.2 runs in the background. Codex starts the run, checks on it and gives you the text, json when it's ready.
What is GLM-5.2 best suited for?
Start with engineering work that spans many steps: implementing a specification, debugging a complex failure, or optimizing an existing system. Its training specifically covers sustained coding tasks.
How does GLM-5.2 differ from GLM-5.1?
GLM-5.2 increases the maximum context from 200K to 1M tokens and adds reasoning effort control. That gives you more room for project context and a way to choose how much reasoning a task receives.
Can GLM-5.2 write a million-token response?
Z.ai specifies a 1M-token context window and a separate 128K-token maximum output. Plan large deliverables in sections that fit the output limit.
Can GLM-5.2 directly analyze images or audio?
Z.ai lists text as both its input and output modality. Supply a text transcript, extracted text, or a written description when the original material uses another format.
When should I choose higher reasoning effort?
High balances reasoning depth and latency. Try Max for difficult debugging, hard math, or plans with several dependent steps, then compare the results on your actual task.
Start using GLM-5.2 in Codex
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Codex:
set up https://spiral.ai/SKILL.md


