Gemini2.5Pro API Returns Free, Developer Community Responds Enthusiastically

AIbase基地

Published inAI News · 5 min read · Jun 30, 2025

Recently, Google announced that the API of its flagship AI model Gemini 2.5 Pro has been re-added to the free tier of Google AI Studio, a move that has sparked widespread attention and heated discussions within the developer community. According to AIbase, this initiative marks another significant step in Google's efforts to popularize AI technology, offering developers more accessible opportunities for innovation.

Gemini 2.5 Pro is Google's most advanced AI model to date, renowned for its exceptional multimodal capabilities and strong reasoning performance. The model supports a context window of up to 1 million tokens and can process various input types, including text, images, audio, and video. It performs particularly well in tasks such as coding, complex reasoning, and data analysis. Previously, in May, Google had adjusted the policies of Google AI Studio, restricting free users from accessing the Gemini 2.5 Pro series models, which raised concerns among some developers about the cost of accessing high-performance models. Now, the return of the free API tier is undoubtedly a positive response to developers' needs.

According to AIbase, a representative from Google's AI API platform stated: "We are committed to providing free API packages to millions of developers. Gemini 2.5 Pro is now back in the free tier and can be used for free in the Gemini app. To meet high demand, we have doubled the free computing resources and are exploring potential paths to continue opening up the free version in the future, such as setting lifetime call limits or launching incentive programs." This statement highlights Google's efforts to balance the commercialization of the model with user experience.

Developer communities have reacted positively to this news. AIbase noted that the return of Gemini 2.5 Pro to the free tier provides developers with opportunities to conduct high-performance testing and prototype development on Google AI Studio and Vertex AI, especially in scenarios requiring handling large codebases or multimodal tasks. Its leading performance on the WebDev Arena and LMArena rankings further solidifies its position in the AI development field. Compared to other competitive models, such as OpenAI's o3-mini and DeepSeek's R1, Gemini 2.5 Pro demonstrates unique advantages in multimodal processing and reasoning capabilities, although its API pricing is slightly higher than some competitors when dealing with large input volumes.

However, AIbase also mentioned that the free-tier Gemini 2.5 Pro still has limitations in terms of lower call rates. Developers who need higher performance and greater data processing capabilities may still consider paying tiers. According to Google's previous pricing announcement, the input price for the Gemini 2.5 Pro API is $1.25 per million tokens (approximately 9.1 RMB), and the output price is $10 per million tokens (approximately 72.9 RMB). After exceeding 200,000 tokens, the prices increase to $2.5 and $15 respectively. Despite this, the return of the free tier provides valuable opportunities for budget-constrained developers to explore cutting-edge AI technologies.

AI Daily Report - June 30th: Baidu Open Sources the WENXIN Large Model 4.5 Series; Tongyi Qianwen Multimodal Generation Model Qwen VLo

Welcome to the AIbase [AI Daily Report] section! Spend three minutes a day to learn about the latest AI events, helping you understand AI industry trends and innovative AI product applications. For more AI news, visit: https://www.aibase.com/zh1. Baidu officially releases the WENXIN Large Model 4.5 series and fully opens it to the public, featuring ten new models with various parameter configurations. These models are trained and inferred using the PaddlePaddle framework, achieving a FLOPs utilization rate of 47%, and perform well in multi-modal text tasks.

Test Article

The internal testing project of Xiaomi, "AI Toolkit," has officially announced the end of its phased testing and plans to suspend service starting July 5, 2025. As an important AI project incubated internally by Xiaomi, the AI Toolkit aims to explore and integrate cutting-edge AI technologies, providing users with a series of innovative features and experiences. Although the specific testing functions and application scenarios have not been fully disclosed, its name suggests its positioning as a multifunctional AI toolset. During the recent testing period, the AI Toolkit has gathered some Xiaomi employees

Test Article

The internal testing project of Xiaomi, "AI Toolbox," has officially announced the end of its phased internal testing and plans to suspend services starting from July 5, 2025. As an important AI project incubated internally by Xiaomi, the AI Toolbox aims to explore and integrate cutting-edge AI technologies, providing users with a series of innovative features and experiences. Although the specific internal testing functions and application scenarios have not been fully disclosed, its name suggests its positioning as a multifunctional AI toolkit. During the recent internal testing period, the AI Toolbox has gathered some Xiaomi employees

Baidu Launches the WENXIN Large Model 4.5 Series Open Source, Sparking a New Wave in the Domestic Large Model Market!

Recently, Baidu officially announced the open-source release of its WENXIN Large Model 4.5 series, launching a total of ten models, including mixed expert (MoE) models with 47B and 3B activated parameters, as well as dense models with 0.3B parameters. This open-source initiative not only fully publicizes the pre-trained weights but also provides inference code, marking a significant advancement for Baidu in the field of large models. These newly released models can be downloaded and deployed on platforms such as PaddlePaddle Starry Sky Community and Hugging Face. Additionally, Baidu Intelligent Cloud's Qianfan Large Model Platform also provides

The Internal Testing Period of Xiaomi AI Toolbox Ends, Service Will Be Suspended Starting July 5

The internal testing project "Xiaomi AI Toolbox" has officially announced the end of its phased internal testing and plans to suspend service starting July 5, 2025. "AI Toolbox" is an important AI project incubated internally by Xiaomi, aimed at exploring and integrating cutting-edge AI technologies to provide users with a series of innovative features and experiences. Although the specific internal testing functions and application scenarios have not been fully disclosed, its name suggests its positioning as a multifunctional AI toolset. During the recent internal testing period, "AI Toolbox" has gathered some Xiaomi employees and core users.

The 'In-Depth Research' Feature of Doubao is Now in Testing on the Doubao APP, Web Version, and Desktop Version

Recently, the Doubao APP, web version, and desktop version platforms have introduced a new feature test - the 'In-Depth Research' feature has been officially launched, offering users free trial. This feature aims to help users efficiently handle complex tasks by quickly integrating massive in-depth information and generating detailed research reports or visualized web results.