Models & Providers

CoStrict supports 75+ LLM providers via AI SDK, plus the built-in CoStrict platform provider.

---

CoStrict Platform

The CoStrict platform is the recommended default provider. After logging in, it is available immediately with no additional API key configuration.

See Account Login to get started.

---

Built-in Providers

The following major providers are built in and ready to use once API keys are configured:

Provider	Provider ID	Get API Key
Anthropic	anthropic	console.anthropic.com
OpenAI	openai	platform.openai.com
Google Gemini	google	aistudio.google.com
Azure OpenAI	azure	Azure Portal
Amazon Bedrock	amazon-bedrock	AWS Console
Google Vertex AI	google-vertex	Google Cloud Console
Groq	groq	console.groq.com
Mistral	mistral	console.mistral.ai
DeepSeek	deepseek	platform.deepseek.com
xAI (Grok)	xai	console.x.ai
Ollama (local)	ollama	No key needed
LM Studio (local)	lmstudio	No key needed

Additional providers can be added via Custom Providers.

---

Configuring API Keys

Via CLI

Use cs auth login to configure API keys interactively. Keys are stored in ~/.local/share/costrict/auth.json:

cs auth login

Via Environment Variables

CoStrict automatically reads API keys from environment variables on startup:

export ANTHROPIC_API_KEY=sk-ant-...
export OPENAI_API_KEY=sk-...
export GOOGLE_GENERATIVE_AI_API_KEY=...

You can also define them in a .env file in the project root.

---

Selecting a Model

Type /models in the TUI to open the model selection UI, or use the shortcut <leader>m.

Specify a model on the command line:

cs --model anthropic/claude-sonnet-4-5
cs run --model costrict/GLM-4.7 "Hello"

---

Recommended Models

The following models perform well for code generation and tool use:

• Claude Opus 4.5
• Claude Sonnet 4.5
• GPT-5
• Gemini 3 Pro

---

Setting the Default Model

Set the model field in the config file using the format provider_id/model_id:

costrict.json

{
  "$schema": "https://opencode.ai/config.json",
  "model": "anthropic/claude-sonnet-4-5"
}

Use small_model for lightweight background tasks like title generation:

costrict.json

{
  "$schema": "https://opencode.ai/config.json",
  "model": "anthropic/claude-sonnet-4-5",
  "small_model": "anthropic/claude-haiku-4-5"
}

Model Loading Priority

1. --model / -m CLI flag
2. model field in config file
3. Last used model
4. First available model by internal priority

---

Configuring Model Options

Set global options for a specific model:

costrict.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "anthropic": {
      "models": {
        "claude-sonnet-4-5-20250929": {
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 16000
            }
          }
        }
      }
    },
    "openai": {
      "models": {
        "gpt-5": {
          "options": {
            "reasoningEffort": "high"
          }
        }
      }
    }
  }
}

---

Model Variants

Many models support multiple variants with different configurations. Use ctrl+t (variant_cycle) to switch between them.

Built-in Variants

Anthropic:

• high — High thinking budget (default)
• max — Maximum thinking budget

OpenAI (varies by model):

• none / minimal / low / medium / high / xhigh — Reasoning intensity levels

Google:

• low — Lower reasoning/token budget
• high — Higher reasoning/token budget

Custom Variants

costrict.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "openai": {
      "models": {
        "gpt-5": {
          "variants": {
            "thinking": {
              "reasoningEffort": "high",
              "textVerbosity": "low"
            },
            "fast": {
              "disabled": true
            }
          }
        }
      }
    }
  }
}

---

Custom Providers

Connect any provider that exposes an OpenAI-compatible API:

costrict.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "myprovider": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "My Provider",
      "options": {
        "baseURL": "https://api.myprovider.com/v1",
        "apiKey": "{env:MY_API_KEY}"
      },
      "models": {
        "my-model": {
          "name": "My Model",
          "limit": {
            "context": 200000,
            "output": 65536
          }
        }
      }
    }
  }
}

Config file locations:

• Project-level: costrict.json in the project root
• Global: ~/.config/costrict/costrict.json

---

Amazon Bedrock

Amazon Bedrock supports additional AWS-specific configuration:

costrict.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "amazon-bedrock": {
      "options": {
        "region": "us-east-1",
        "profile": "my-aws-profile"
      }
    }
  }
}

• region — AWS region (defaults to AWS_REGION env var)
• profile — Named AWS profile from ~/.aws/credentials

---

Listing Available Models

cs models

# Filter by provider
cs models anthropic

# Refresh the model cache
cs models --refresh