Skip to main content

​
Overview

ScryCLI supports multiple AI providers and models for code analysis. You can select your preferred model through an interactive CLI interface or by manually editing the configuration file.

​
Available Providers

ScryCLI supports the following AI providers through OpenRouter:
ProviderModels AvailableRecommended Model
OpenAI2 modelsgpt-oss-120b
ArceeAI2 modelsTrinity Large
Qwen2 modelsQwen3 Next 80B A3B Instruct
Mistral1 modelMistral Small 3.1 24B
StepFun1 modelStep 3.5 Flash
Zai1 modelGLM 4.5 Air
All models listed are free-tier models accessible through OpenRouter with rate limits.

​
Selecting a Model

​
Using the CLI Command

1

Open model selection

While ScryCLI is running, type:
/model
2

Choose your provider

Use arrow keys to navigate and Enter to select:
> OpenAI
  StepFun
  Qwen
  ArceeAI
  Mistral
  Zai
3

Select a model

After choosing a provider, select from available models:
> gpt-oss-120b (Recommended)
  gpt-oss-20b
4

Configuration saved

Your selection is automatically saved to ~/.scrycli/config.json

​
Available Models

​
OpenAI Models

// From src/ui/SelectModel.tsx:21-24
openai: [
  { label: "gpt-oss-120b (Recommended)", value: "openai/gpt-oss-120b:free" },
  { label: "gpt-oss-20b", value: "openai/gpt-oss-20b:free" },
]
Recommended: openai/gpt-oss-120b:free - Best balance of capability and availability

​
ArceeAI Models

arceeai: [
  { label: "Trinity Large (Recommended)", value: "arcee-ai/trinity-large-preview:free" },
  { label: "Trinity Mini", value: "arcee-ai/trinity-mini:free" },
]
Recommended: arcee-ai/trinity-large-preview:free - Optimized for code tasks

​
Qwen Models

qwen: [
  { label: "Qwen3 Coder 480B A35B", value: "qwen/qwen3-coder:free" },
  { label: "Qwen3 Next 80B A3B Instruct (Recommended)", value: "qwen/qwen3-next-80b-a3b-instruct:free" },
]
Recommended: qwen/qwen3-next-80b-a3b-instruct:free - Excellent instruction following

​
Mistral Models

mistral: [
  { label: "Mistral Small 3.1 24B", value: "mistralai/mistral-small-3.1-24b-instruct:free" },
]
Single option: mistralai/mistral-small-3.1-24b-instruct:free

​
StepFun Models

stepfun: [
  { label: "Step 3.5 Flash", value: "stepfun/step-3.5-flash:free" },
]
Single option: stepfun/step-3.5-flash:free - Fast inference

​
Zai Models

zai: [
  { label: "GLM 4.5 Air", value: "z-ai/glm-4.5-air:free" },
]
Single option: z-ai/glm-4.5-air:free

​
Configuration Structure

When you select a model, ScryCLI updates your configuration: 
{
  "model": {
    "modelProvider": "openai",
    "modelName": "openai/gpt-oss-120b:free",
    "modelKey": "your-api-key"
  }
}
The configuration is set via: 
// From src/ui/SelectModel.tsx:68
setConfig('model', {
  modelProvider: selectedProvider,
  modelName: item.value
});

​
Configuration Fields

FieldDescriptionExample
modelProviderProvider identifier"openai", "qwen", "arceeai"
modelNameFull model path"openai/gpt-oss-120b:free"
modelKeyYour API key"sk-or-v1-..."
You must set modelKey separately. The model selection command only sets modelProvider and modelName.

​
Manual Configuration

You can manually edit your model configuration:
1

Open config file

nano ~/.scrycli/config.json
2

Update model settings

{
  "model": {
    "modelProvider": "qwen",
    "modelName": "qwen/qwen3-next-80b-a3b-instruct:free",
    "modelKey": "sk-or-v1-your-key-here"
  }
}
3

Save and restart

Changes take effect when you restart ScryCLI

​
Model Selection Process

The model selection UI follows a two-step process: 
// From src/ui/SelectModel.tsx:44-75
const SelectModel = (props: selectModelType) => {
    const [selectedProvider, setSelectedProvider] = useState<string | null>(null);
    const [selectedModel, setSelectedModel] = useState<string | null>(null);

    // Step 1: Select provider
    if (!selectedProvider) {
        return <SelectInput items={providers} onSelect={setSelectedProvider} />;
    }

    // Step 2: Select model from provider's available models
    if (selectedProvider && !selectedModel) {
        return <SelectInput 
            items={modelsMap[selectedProvider]} 
            onSelect={(item) => {
                setSelectedModel(item.value);
                setConfig('model', { 
                    modelProvider: selectedProvider, 
                    modelName: item.value 
                });
            }} 
        />;
    }
};

​
Viewing Current Model

Your currently selected model is displayed in the ScryCLI interface: 
// From src/ui/InputBox.tsx:40
<Text color='yellow'>
  {`Model: ${config?.model?.modelName ?? 'Not Selected'}`}
</Text>
If no model is selected, you’ll see “Model: Not Selected”

​
Model Usage

The selected model is used when making AI calls: 
// From src/model/openRouter.ts:19-23
const result = openRouterClient.callModel({
  model: `${getConfig().model.modelName}`,
  instructions: systemPrompt,
  input: `${prompt} \n\nFile Tree: ${fileTreeString}`,
});

​
Validation Requirements

Before ScryCLI can make model calls, it validates: 
// From src/lib/isModelSelected.ts:6-15
const isModelSelected = (): boolean => {
    const config = getConfig();
    if (!config?.model?.modelName) return false;
    if (!config?.model?.modelKey) return false;
    return config.model.modelName !== '' && config.model.modelKey !== '';
};
Both modelName and modelKey must be:
  • Present in the configuration
  • Non-empty strings

​
Choosing the Right Model

​
For Code Analysis

Recommended: qwen/qwen3-coder:free
  • Specialized for code understanding
  • Excellent at parsing syntax
  • Good for large codebases

​
For General Tasks

Recommended: openai/gpt-oss-120b:free
  • Balanced performance
  • Good instruction following
  • Widely available

​
For Speed

Recommended: stepfun/step-3.5-flash:free
  • Fastest inference
  • Lower latency
  • Good for quick queries

​
Troubleshooting

​
Model Not Selected Error

If you see “Model: Not Selected”:
1

Run model selection

/model
2

Verify configuration

cat ~/.scrycli/config.json
Ensure model.modelName exists
3

Add API key

Make sure model.modelKey is also set

​
Model Calls Failing

If model calls fail:
  1. Check model name format
    • Must be in format: provider/model:tier
    • Example: openai/gpt-oss-120b:free
  2. Verify API key
  3. Check OpenRouter status
    • Some models may be temporarily unavailable

​
Next Steps

API Keys

Configure authentication for model access

Configuration Setup

Learn about the configuration system