GLM-5.3-Flash in OpenCode: Free Setup Guide
GLM-5.3-Flash is free to use inside OpenCode for a limited time as the model opencode/glm-5.3-flash-free (OpenCode Zen lists it at $0 for input, output and cached reads, and says its provider keeps zero data retention and doesn't train on your data). Install OpenCode, run /models and pick GLM-5.3-Flash Free, or set "model": "opencode/glm-5.3-flash-free" in opencode.json. The paid version, glm-5.3-flash, costs $0.15 per 1M input tokens and $0.50 per 1M output tokens on OpenCode Zen. GLM-5.3-Flash is Z.ai's multimodal coding model with 320B total / 18B active parameters and a 1M-token context.
GLM-5.3-Flash in OpenCode at a glance
| Free in OpenCode | Paid (OpenCode Zen) | |
|---|---|---|
| Model id | opencode/glm-5.3-flash-free |
opencode/glm-5.3-flash |
| Price per 1M tokens | Free (limited time) | $0.15 input · $0.50 output · $0.03 cached read |
| Data | Provider follows zero-retention, no training (per OpenCode) | See OpenCode Zen privacy section |
| Best for | Trying the model in the OpenCode app | API use, other tools, steady workloads |
Source: OpenCode Zen docs, checked October 18, 2026. "Limited time" has no end date, so check the docs before you build a workflow on the free model.
How to use GLM-5.3-Flash for free in OpenCode
- Install OpenCode (from the official site):
curl -fsSL https://opencode.ai/install | bash
# or
npm i -g opencode-ai
- Start OpenCode in your project folder:
opencode - Type
/modelsand choose GLM-5.3-Flash Free. - Don't see it? Refresh the model list:
opencode models opencode --refresh | grep free
Make it the default model (opencode.json)
{
"$schema": "https://opencode.ai/config.json",
"model": "opencode/glm-5.3-flash-free"
}
One-off run from the terminal
opencode run -m opencode/glm-5.3-flash-free "Explain what this repo does"
What GLM-5.3-Flash is
From Z.ai's model docs:
| GLM-5.3-Flash | |
|---|---|
| Maker | Z.ai (Zhipu) |
| Size | 320B total parameters, 18B active (mixture of experts) |
| Context | 1M tokens |
| Input | Text, images, video and files (native multimodal) |
| Thinking | Always on (thinking.type only supports enabled) |
| Recommended settings | temperature: 1, top_p: 0.95, reasoning_effort: max |
| Faster variant | GLM-5.3-FlashX |
| Also on | Z.ai API and the GLM Coding Plan (3× the quota of GLM-5.3) |
GLM-5.3-Flash vs GLM-5.3 price (OpenCode Zen, per 1M tokens)
| Model | Input | Output | Cached read |
|---|---|---|---|
| GLM-5.3-Flash Free | Free | Free | Free |
| GLM-5.3-Flash | $0.15 | $0.50 | $0.03 |
| GLM-5.3 | $1.40 | $4.40 | $0.26 |
The Flash model costs about a tenth of full GLM-5.3 on Zen.
Limits to know
- Use it inside OpenCode. The free model is meant for the OpenCode app. Calling the free id from other tools through the raw endpoint has been reported to fail with a free-tier error. For API work, use the paid
glm-5.3-flash. - Thinking can't be turned off, so simple answers take longer than with a non-reasoning model.
- Free promotions end. OpenCode doesn't give an end date. Keep a paid fallback in your config.
Other free models in OpenCode right now
OpenCode Zen also lists DeepSeek V4.1 Flash Free (deepseek-v4.1-flash-free) as free for a limited time. See our DeepSeek V4.1-Flash guide for that model's specs and API prices.
Frequently asked questions
Is GLM-5.3-Flash free in OpenCode?
Yes, for a limited time. OpenCode Zen lists GLM-5.3-Flash Free (opencode/glm-5.3-flash-free) at $0 for input, output and cached reads. No end date is given.
How do I use GLM-5.3-Flash in OpenCode?
Install OpenCode (curl -fsSL https://opencode.ai/install | bash or npm i -g opencode-ai), start it, type /models and choose GLM-5.3-Flash Free. Or set "model": "opencode/glm-5.3-flash-free" in opencode.json.
What is the model id for GLM-5.3-Flash in OpenCode?
opencode/glm-5.3-flash-free for the free version and opencode/glm-5.3-flash for the paid version.
How much does GLM-5.3-Flash cost?
On OpenCode Zen the paid glm-5.3-flash costs $0.15 per 1M input tokens, $0.50 per 1M output tokens and $0.03 per 1M cached-read tokens. Full GLM-5.3 costs $1.40 / $4.40 / $0.26.
Does OpenCode keep my data when I use the free GLM-5.3-Flash?
OpenCode says the provider of GLM-5.3-Flash Free follows a zero-retention policy and does not use your data for model training.
Can I use opencode/glm-5.3-flash-free outside OpenCode?
It is meant for the OpenCode app. Calls from other tools have been reported to fail with a free-tier error, so use the paid glm-5.3-flash for API work.
What is GLM-5.3-Flash?
Z.ai's native multimodal coding model with 320B total and 18B active parameters, a 1M-token context and always-on thinking. A faster variant, GLM-5.3-FlashX, also exists.
Why doesn't GLM-5.3-Flash Free show in my OpenCode model list?
Refresh the list with opencode models opencode --refresh, update OpenCode, then run /models again.