Google Gemini 1.5 Pro 002 & Gemini 1.5 Flash 002 are now available on Bollwerk AI.

The new models are available to alpha users.

What's New?

Google has introduced two updated models:

1. Gemini-1.5-Pro-002

2. Gemini-1.5-Flash-002

These models build upon the experimental releases showcased at Google I/O in May, offering substantial enhancements across various AI tasks.

Key Highlights

- Reduced Pricing: Google has slashed the price of Gemini 1.5 Pro by over 50% for both input and output on prompts under 128K tokens.

- Increased Rate Limits: The rate limits have been doubled for Gemini 1.5 Flash (now 2,000 RPM) and tripled for Gemini 1.5 Pro (now 1,000 RPM).

- Improved Performance: Both speed and efficiency have been enhanced, with 2x faster output and 3x lower latency.

- Updated Default Filters: Google has refined their content filters for better results.

Performance Improvements

The new Gemini models show impressive gains in several areas:

- 7% increase in MMLU-Pro benchmark

- 20% improvement in MATH and HiddenMath benchmarks

- 2-7% enhancement in vision and code-related tasks

Looking Ahead

Google has hinted at further rate limit increases in the coming weeks, which could unlock even more potential for AI development. At Bollwerk AI, we're committed to staying at the forefront of these advancements and integrating them into our offerings to provide you with cutting-edge AI solutions.