Google Releases Gemini 4 Hydrogen Model with 2.5x Speed Improvement
Google has officially released the Gemini 4 Hydrogen model, focusing on efficient coding and AI agent tasks. Google says that in the default High thinking mode, its task completion speed is 2.5 times that of Gemini 3.8 Flash, and it scores 57% on the Terminal-Bench 4.1 coding benchmark, compared with 18.2% for the previous generation. The model supports a context window of up to 1 million tokens and has been launched globally through the Gemini app, Google Search, Google AI Studio, and the Gemini API. API pricing is $1.5 per million input tokens and $7.5 per million output tokens.
Why read this
SelectedAI score 87
The original text lists comparisons with the previous generation in speed, coding benchmarks, and token consumption, which can be used to assess the positioning of this efficient execution model.
Topics
Body · AI translation
New crowdsourced testing promotional copy New crowdsourced testing promotional copy end Add App-exclusive traffic redirect
Fast Technology News, November 13: Today, Google officially released its all-new AI model Gemini 4 Hydrogen, focused on efficient programming and AI agent tasks, with major upgrades in running speed, coding capability, and reliability.
Google said that compared with the previous generation Gemini 3.8 Flash, Gemini 4 Hydrogen in the default High thinking mode completes tasks 2.5 times faster than the former, and its programming test scores have risen to more than 3 times.
Specifically, in the Terminal-Bench 4.1 programming benchmark, Gemini 4 Hydrogen scored 57%, while Gemini 3.8 Flash scored only 18.2%, a very notable improvement.
At the same time, the new model generates fabricated content more than 5 times less often, and can complete tasks more stably even when tool calls fail or search results are empty.
In terms of programming capability, Gemini 4 Hydrogen can handle tasks such as code writing, program debugging, and vulnerability fixing, and can make modifications targeting specific issues in a codebase, avoiding impact on unrelated code.
Google has also optimized for large codebases, supporting parallel file reading and code search to further improve development efficiency for complex projects.
According to early tests, in complex charts, SQL, and data-intensive enterprise tasks, the new model runs 2 to 2.8 times faster than the previous generation while reducing Token consumption by 12% to 23%.
It is worth mentioning that although Gemini 4 Hydrogen is positioned as an efficient execution model, it has already surpassed Google's own high-end models in some tests.
In the Artificial Analysis Intelligence Index test, its default High mode score reached 51 points, exceeding Gemini 4 Argon's 50 points, and after enabling the highest xhigh thinking mode, the score can rise to 54 points.
Gemini 4 Hydrogen supports a context window of up to 1 million Tokens and can handle multiple types of content such as text, images, audio, video, and complex documents.
Currently, Gemini 4 Hydrogen has begun rolling out to users worldwide, covering platforms including the Gemini app, Google Search, Google AI Studio, and the Gemini API, and free users also have a chance to experience it.
Among them, free users of the Gemini app can use the model through an automatic model selection mechanism, while developers can call it through the Gemini API, with an input price of $1.5 per million Tokens and an output price of $7.5 per million Tokens.
[This article ends] If you need to reprint, please be sure to indicate the source: Fast Technology
Editor in charge: Shi Qi article_adlist[
Report article content ]article_adlist Statement: This is an exclusive article of Sina.com, and unauthorized reproduction is prohibited. tech_web_article_v2015_block_bottom Non-targeted 300*250 button 17/09 wenjing begin
Non-targeted 300*250 button end