# Critical Change: API Key Transmission Method Updated API key transmission has been updated for improved security and compatibility. Previously, the API key was sent as a URL query parameter (e.g., `?key=YOUR_API_KEY`). Now, all functions send the API key using the `HTTP header x-goog-api-key`, following Google Gemini API best practices. This change enhances both security and future compatibility. Users can continue to set their API key using setAPI("YOUR_API_KEY"). There is no change to the user-facing function interfaces. The old method of passing the API key as a URL parameter is no longer supported. See issue #52 ## What's Changed * Add figure support to gemini_narrative function by @jhk0530 in https://github.com/jhk0530/gemini.R/pull/46 * feat: Add gemini_garden function by @jhk0530 in https://github.com/jhk0530/gemini.R/pull/47 * Update gemini_docs.R by @jhk0530 in https://github.com/jhk0530/gemini.R/pull/51 **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.15.0...v0.16.0

## What's Changed * Add gemini_narrative function for table descriptions by @jhk0530 in https://github.com/jhk0530/gemini.R/pull/43 * Add gemini_structured function for structured API responses by @jhk0530 in https://github.com/jhk0530/gemini.R/pull/44 **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.14.1...v0.15.0

## What's Changed - Relaxed the `model` parameter validation to support newly released models like `gemini-2.5-flash-preview-04-17` ### Internal changes - Enhanced error handling across all functions for improved stability - Standardized response processing across all API functions - Updated documentation to reflect expanded model compatibility **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.12.0...v0.13.0

## What's Changed - Added Gemini searching `gemini_search` for ground using gemini-2.0 models - Added Gemini searching `gemini_searchR` for retrive using gemini-1.5 models - styler applied to package. **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.11.0...v0.12.0

## What's Changed - Added Gemini 2.5-pro-exp model (`2.5-pro-exp-03-25`) - Deprecated Gemini 2.0-pro-exp model (`2.0-pro-exp-02-05`) - Now image generation possible with `gen_image()` using `2.0-flash-exp-image-generation` model. **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.10.0...v0.11.0

## What's Changed - `countTokens` function for countTokens API - `setEnv` function added. - Add CRANlogs badge in readme ### Important - Deprecated Gemini 1.5 models (`1.5-pro`, `1.5-flash`) **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.9.0...v0.10.0

## What's Changed - Updated parameters across all functions (`temperature`, `maxOutputTokens`, `topK`, `topP`, `seed`) - Adjusted default values: temperature to **1** and maxOutputTokens to **8192**, model as **2.0-flash** ### Important - {gemini.R} now only supports model for "2.0-flash", "2.0-flash-lite", "1.5-flash", "1.5-flash-8b", "1.5-pro". (deprecated '1.0-pro' and '2.0-flash-exp') **Full Changelog**: https://github.com/jhk0530/gemini.R/compare/v0.8.0...v0.9.0

## What's Changed - Add Vertex AI with new functions: `token.vertex()`, `gemini.vertex()`, `gemini_image.vertex()`, `gemini_audio.vertex()` - These functions are in experimental stage and may change in future versions.