Z.ai · checked October 13, 2026
How do I run GLM 5.3 Flash in OpenCode, Claude Code or Codex?
GLM 5.3 Flash runs on Standard Compute at https://api.stdcmpt.com/v1 with the model ID z-ai/glm-5.3-flash. The same key works in OpenCode, Claude Code and Codex.
| Model | The fast, low-cost member of Z.ai's GLM 5.3 coding release. Unlike GLM 5.2 and GLM 5.3, it reads images. |
|---|---|
| Model ID | z-ai/glm-5.3-flashShort form glm-5.3-flash also works |
| Base URL | https://api.stdcmpt.com/v1Claude Code uses https://api.stdcmpt.com |
| Context | 1,048,576 tokens; about 850K input and 50K output per request |
| Input | Text and images |
| Reasoning | Always on; effort minimal to high |
Copy a working config
OpenCode is where most people run GLM. The same model ID works in Claude Code and Codex through the launcher.
{
"$schema": "https://opencode.ai/config.json",
"model": "standardcompute/z-ai/glm-5.3-flash",
"provider": {
"standardcompute": {
"npm": "@ai-sdk/openai-compatible",
"name": "Standard Compute",
"options": {
"baseURL": "https://api.stdcmpt.com/v1",
"apiKey": "{env:STANDARDCOMPUTE_API_KEY}"
},
"models": {
"z-ai/glm-5.3-flash": { "name": "GLM 5.3 Flash" },
"standardcompute": { "name": "Smart routing" }
}
}
}
}
Replace YOUR_API_KEY with a key from the dashboard; OpenCode reads it from STANDARDCOMPUTE_API_KEY. Full setup: OpenCode, Claude Code, Codex CLI.
What it costs
| Per 1M tokens | GLM 5.3 Flash | GLM 5.3 |
|---|---|---|
| Input | $0.15 | $1.40 |
| Cached input | $0.03 | $0.26 |
| Output | $0.50 | $4.40 |
Z.ai API rates. Cache storage is listed as free for a limited time. Source: Z.ai API pricing, checked October 13, 2026.
GLM 5.3 Flash costs about a ninth of GLM 5.3 per input token and about a twenty-seventh of Claude Opus 5.5. That makes it a strong efficient rung in a routed selection: put it next to a frontier model on the Models page and routine steps land on Flash while hard ones go up.
What works differently here
- Thinking cannot be switched off. The endpoint rejects requests that disable reasoning, so the gateway uses a low effort instead, and in our tests the effort level barely changed how much it thought.
- Image input was tested on the provider we route to, including questions about what the image shows. Video is advertised elsewhere for this model but is not offered here.
- If GLM is the only model you use, compare Z.ai's own GLM Coding Plan first. A Standard Compute plan makes sense when GLM shares the month with other models.
Other ways to run GLM 5.3 Flash
| Z.ai API $0.15 / $0.50 per 1M | First-party, pay per token, with cached input at $0.03. |
|---|---|
| GLM Coding Plan (Z.ai) Subscription | Z.ai's own plan for coding tools. GLM models only, metered in prompt windows. |
| OpenRouter hosts Pay per token | Several hosts serve it; check each one's rate and whether it accepts images. |
| Standard Compute Monthly plan | GLM 5.3 Flash next to Claude, GPT and other open models on one key. |
FAQ
Can GLM 5.3 Flash read screenshots?
Yes. It accepts images on Standard Compute; we checked that it answers questions about image content on the provider we route to. GLM 5.2 and GLM 5.3 are text-only.
GLM 5.3 Flash or GLM 5.3?
Flash for routine edits, tool calls and high-volume work at about a ninth of the input price. GLM 5.3 for harder multi-file reasoning. Select both and routing chooses per request.
Plans from $19/mo
One key for GLM 5.3 Flash and the rest of the catalog, with a monthly compute budget you can watch in the dashboard and no overage charges.
Sources: Z.ai API pricing
Related: Run GLM 5.3 · OpenCode pricing · All model IDs