Switch between Claude and GLM APIs instantly. Monitor usage in real-time. Manage multiple sessions with different models.
Everything you need to manage multiple AI providers and monitor your API usage efficiently.
Instantly switch between Anthropic Claude and GLM (Z.AI) APIs without changing your Claude Code configuration.
Track token usage, request counts, and costs in real-time. View stats by model, daily, or session.
Run multiple Claude Code instances with different vendors and models. Each session is independently configurable.
API keys are stored securely in local encrypted storage. Your credentials never leave your machine.
Full support for English, Korean, Japanese, and Chinese interfaces.
Runs quietly in system tray. Quick access to start/stop proxy and view status at any time.
Intercept messages with >>swiftcast to run shell commands, HTTP requests, or read files.
Powerful hooks for logging, context injection, and custom behavior. Configure per-session or system-wide.
SwiftCast acts as a local proxy between Claude Code and your AI providers. See it in action.
Register your Anthropic (Claude) and GLM API accounts. Use Auto Scan on macOS to automatically import existing Claude credentials from Keychain, or add accounts manually with your API keys.
Launch the proxy with one click. SwiftCast automatically configures Claude Code's settings.json to route through the local proxy. Switch between providers instantly.
Run multiple Claude Code instances with different configurations. Assign different vendors and models to each session. Perfect for A/B testing or cost optimization.
Track token consumption across all sessions. View detailed breakdowns by model, day, or session. Optimize costs by understanding your usage patterns.
SwiftCast is perfect for developers who need flexibility with AI providers.
Use different providers for different tasks. Route simple tasks to cheaper models while keeping complex work on Claude Opus.
Compare outputs between Claude and GLM for the same prompts. Run parallel sessions with different models to evaluate quality.
Monitor token consumption across all your AI interactions. Track costs by project using different sessions.
When one provider is rate-limited or down, quickly switch to another without interrupting your workflow.
Extend SwiftCast with powerful automation and customization features.
Intercept Claude messages starting with >>swiftcast <task> to execute shell commands, HTTP requests, or file operations. Configure in ~/.sessioncast/tasks.json.
Log all API requests and responses to JSON files for debugging, auditing, or analysis. Enable/disable per session.
Automatically inject custom context during conversation compaction. Keep your project rules and preferences persistent across long sessions.
Override system settings per session. Enable/disable logging, custom tasks, or context injection for specific sessions via API.
Download SwiftCast and take control of your AI provider management.