# Gemini 2.5 Flash is supposed to bring more affordable AI reasoning to tasks that require more thinking.

*genai · news · 2025-04-18 · BGR*

## Key points

- Gemini 2.5 Flash is Google's first fully hybrid reasoning AI model with togglable reasoning features.
- Developers can set a precise thinking budget from 0 to 24,576 tokens in Google AI Studio and Vertex AI.
- Reasoning-capable output costs up to $3.50 per million tokens, six times standard output rates.
- Gemini 2.5 Flash achieved the second-highest score after ChatGPT o4-mini in Humanity's Last Exam benchmark.
- Gemini 2.5 Flash is Google's fastest, most cost-efficient reasoning model, targeting developers and enterprises.

**Companies:** Google
**Countries:** United States

[Read the full story on BGR](https://bgr.com/tech/gemini-2-5-flash-is-googles-cheapest-thinking-ai-what-you-need-to-know/)

---

Canonical: https://newsio.io/n/7c838dc7-8312-4c47-b774-2fd6dbcbce56/gemini-2-5-flash-is-supposed-to-bring-more-affordable-ai-reasoning-to-tasks-that
Summarized by Newsio from BGR. https://newsio.io/how-it-works