Free Models
Best for: You want to complete your first OpenCode conversation with 0 cost and 0 configuration.
OpenCode includes a set of completely free models that require no registration or API Key - they work immediately after installation.
Current Free Models
| Model ID | Name | Features |
|---|---|---|
glm-4.7-free | GLM-4.7 | By Zhipu, supports reasoning, excellent Chinese performance |
minimax-m2.1-free | MiniMax M2.1 | By MiniMax, supports reasoning, 200K context |
gpt-5-nano | GPT-5 Nano | By OpenAI, lightweight and fast |
kimi-k2.5-free | Kimi K2.5 | By Moonshot, supports reasoning, 256K context |
big-pickle | Big Pickle | Hidden model, limited-time free trial |
Note
The free model list may change over time. The list shown in /models command is always accurate.
Free models have IP-level usage limits, suitable for trial and light use. For higher quotas, you can log in to OpenCode Zen to get an API Key.
Follow Along
Step 1: Start OpenCode
opencodeOn first launch, OpenCode automatically loads free models with no configuration needed.
Step 2: Select a Free Model
Enter:
/modelsChoose any free model from the list, for example:
opencode/glm-4.7-freeopencode/minimax-m2.1-free
How to identify free models?
Free models usually have a -free suffix, or show $0 price in the /models list.
Step 3: Send Your First Message
Hello, please introduce yourselfIf you receive a response, you've successfully completed the "Install → Select Model → Chat" workflow.
Checklist
- [ ] Running
/modelsshowsopencode/*-freeor models with$0price - [ ] Sending a message receives an AI response
Common Issues
| Issue | Cause | Solution |
|---|---|---|
No free models in /models | Network issue or outdated version | Check network, run opencode upgrade to update |
| Request rejected / rate limited | Free models have IP-level quotas | Try a different free model or wait a while |
| Slow response | Free model servers under high load | Normal behavior, be patient or try another model |
Want More?
Free models are great for trial and light use. If you need:
- Higher request quotas
- More model choices (Claude, GPT-5, etc.)
- More stable service
You can log in to OpenCode Zen to get an API Key. See 1.4 Connect Models for more details.
Next Steps
- Return to 1.4 Overview to choose your next path
- Or proceed to 2.1 Interface & Basics to learn daily usage