Version 1.0.76

What: Added support for Claude Sonnet 4 model with 1-million token context window How to use:

# Switch to the 1M context model through the model selector
# The model will appear as "Sonnet (1M context)" in your model options
# Note: This feature requires API access verification

Details:

• Available as "Sonnet (1M context)" or "Sonnet 4 (with 1M token context)" in the model selector
• Provides 1,000,000 tokens of context (5x the standard 200,000 limit)
• Perfect for working with large codebases, lengthy documents, or extended coding sessions
• Pricing: $6 per million input tokens, $22.50 per million output tokens
• Access is controlled via API verification - only available to eligible accounts
• Uses rate limits faster than standard Sonnet on long requests
• Context warnings are intelligently adjusted to avoid premature alerts when you still have hundreds of thousands of tokens available

What changed: Refactored Node.js module imports to use named imports instead of default imports Impact: Reduces memory footprint and improves code clarity by explicitly importing only the required functions from crypto, stream, and process modules

What changed: Context warnings now adapt based on the model's maximum context size Impact: Users with 1M context models won't receive premature "low context" warnings when they still have substantial tokens remaining, improving the experience for long-form conversations

No bug fixes were included in this release.