How to Use GLM 5.2 (batch) in Muse Code
Use GLM 5.2 for complex debugging, code migrations, and research that needs substantial context. It expands on GLM 5.1 with a 1M-token context for sustained work across large projects.
Prompts to try with GLM 5.2 (batch) in Muse Code
Copy an example into Muse Code, or ask in your own words.
Use GLM 5.2 to review this repository and map its modules, dependencies, API contracts, and main data flows. Identify three areas where a change could affect unrelated behavior, citing the relevant files.
Use GLM 5.2 to migrate this project's deprecated SDK calls to the supported equivalents. First identify affected callers and compatibility risks, then implement the migration and run the relevant tests.
Use GLM 5.2 to investigate the failing integration test. Trace the behavior across the relevant files, reproduce the failure, make a targeted fix, and report the verification results.
Use GLM 5.2 with greater thinking effort to investigate the slowest path in this project's profiling results. Establish a baseline, test one proposed optimization, and compare correctness and performance.
Use GLM 5.2 to decouple this project's authentication module from its storage implementation. Preserve public API signatures and runtime behavior, explain the impact scope, and verify the change.
How to set up GLM 5.2 (batch) in Muse Code
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Install Muse Code
curl -fsSL https://dev.meta.ai/install.sh | bash
- Ask Muse to add the server
Start muse and ask:
Add the Spiral MCP server at https://api.spiral.ai/mcp
- Sign in
In your terminal, run:
muse mcp login spiral
Use the Spiral skill
- Send Muse Code 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 (batch)
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:batch", "input": { "messages": [ { "role": "user", "content": "Hello!" } ] } }'
Then paste one of the prompts above: Muse Code finds GLM 5.2 (batch) on Spiral, shows the price and runs it.
GLM 5.2 (batch) providers and prices
Spiral runs GLM 5.2 (batch) through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.005282 per run at the default inputs. $0.4186 per 1M input tokens, $1.2857 per 1M output tokens.
Settings you can ask for
Ask Muse Code 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. On a reasoning model the reasoning tokens count toward it: a cap too low for the effort can end with no answer (finish_reason length). | |
| 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 |
Every API field of Z.ai: GLM 5.2 (batch) is on its tool page: open its card above.
What can you do with GLM 5.2 (batch) in Muse Code?
GLM 5.2 can carry substantial project context through debugging, migrations, and performance work. Give the agent clear boundaries and require evidence that the requested behavior works.
Repository architecture review
Map module responsibilities, dependencies, and API boundaries before planning a change to a large codebase.
API and SDK migrations
Plan and implement changes across callers, adapters, and tests while preserving specified behavior.
Complex bug investigations
Compare code, logs, and failure reports to develop a testable explanation and a targeted fix.
Performance optimization
Use profiling evidence to propose changes, run comparisons, and report whether the measured bottleneck improved.
Research synthesis
Compare lengthy source material, identify disagreements, and produce a brief with evidence for each conclusion.
Structured document extraction
Turn supplied text into JSON records with consistent categories, supporting excerpts, and explicit missing values.
More on GLM 5.2 (batch) and Muse Code
Overview
GLM 5.2 (batch) in other agents
More Muse Code guides
Questions about GLM 5.2 (batch) in Muse Code
Can Muse Code use GLM 5.2 (batch)?
Yes. Add Spiral to Muse Code as an MCP server, and Muse Code can find and run GLM 5.2 (batch) like any built-in tool.
How much does GLM 5.2 (batch) cost in Muse Code?
Up to $0.005282 per run at the default inputs. $0.4186 per 1M input tokens, $1.2857 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Muse Code plan is separate.
Do I need an OpenRouter account to use GLM 5.2 (batch) in Muse Code?
No. Spiral runs GLM 5.2 (batch) 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 (batch) in Muse Code?
About two minutes: install Muse Code, then ask Muse to add the server, then sign in.
Does Muse Code wait while GLM 5.2 (batch) runs?
GLM 5.2 (batch) runs in the background. Muse Code starts the run, checks on it and gives you the text, json when it's ready.
What is GLM 5.2 good at?
GLM 5.2 is suited to engineering tasks that require keeping substantial context across multiple steps, including debugging, migrations, and performance work. It also fits long-document analysis and research synthesis.
How much context and output can GLM 5.2 handle?
Z.ai documents a 1M-token context window and a separate maximum output of 128K tokens. Treat those as the maker's published limits when planning a large task.
Can GLM 5.2 return JSON?
Yes, Z.ai documents structured output support, including JSON. Specify the expected structure and how to represent missing information.
When should I increase thinking effort?
Consider greater thinking effort for difficult debugging or engineering decisions. Z.ai describes effort control as a tradeoff between coding performance and latency; compare results on a representative task.
Where does GLM 5.2 still fall short?
Extended engineering work remains a limitation: Z.ai reports that GLM 5.2 trails Claude Opus 4.8 on SWE-Marathon. For a long implementation, request intermediate checks and explicit verification results.
Start using GLM 5.2 (batch) in Muse Code
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Muse Code:
set up https://spiral.ai/SKILL.md


