Google Gemini 3.8 Flash Launch: New AI Model Could Challenge OpenAI and Anthropic in Coding

Google is reportedly preparing to launch Gemini 3.8 Flash, a coding-focused AI model that could challenge OpenAI and Anthropic with faster performance and improved capabilities.

Google Preparing New Gemini AI Model

Google is reportedly nearing the launch of a new Gemini artificial intelligence model designed to improve its performance in coding and AI-assisted software development.

According to a Source report, the model is internally known as “Skimaki” and could be released as early as September 2, 2026. The model is expected to be called Gemini 3.8 Flash.

The launch could be significant for Google as it attempts to regain momentum in an increasingly competitive AI market dominated by companies such as OpenAI and Anthropic.

Gemini 3.8 Flash Could Focus on Coding

Coding has become one of the most important commercial applications of generative AI. Developers and businesses are increasingly using AI tools to write, debug and modify software, making coding performance an important area of competition among leading AI companies.

According to the report, some Google engineers reportedly preferred Gemini 3.8 Flash over Anthropic’s Opus model during internal testing on Jetski, Google’s coding platform.

If the model performs similarly after its public release, it could strengthen Google’s position in AI-powered coding and software development.

What Is Gemini 3.8 Flash?

Gemini 3.8 Flash is expected to be a smaller and faster model rather than a new flagship frontier AI system.

Google’s Flash models are generally designed to provide quick responses at lower costs compared with larger Pro models. Their relatively smaller size also allows researchers and engineers to experiment with different approaches more efficiently.

This could give Google greater flexibility when developing and testing coding capabilities across multiple teams.

Google Accelerates Flash Model Development

Google has been increasing the pace of its Flash model releases. The company reportedly introduced Gemini 3.6 Flash in July, followed by Gemini 3.7 Flash only a few weeks later.

The rapid development cycle highlights Google’s effort to improve its AI models while keeping pace with competitors.

The company previously appeared to gain momentum after launching Gemini 3.0, but subsequent advances from OpenAI and Anthropic have increased pressure on Google to deliver stronger models.

Delays Affect Gemini’s AI Strategy

Google’s AI development has also faced setbacks. Its next flagship model, Gemini 3.5 Pro, has reportedly experienced delays after candidate versions failed to deliver sufficient improvements on internal coding benchmarks.

According to the report, some internal versions were abandoned because they did not provide a significant enough advantage over Google’s Flash models.

Meanwhile, Google is also developing Gemini 4, its next major flagship AI model. The model has reportedly performed well during pretraining evaluations but is still undergoing post-training work.

Leadership Changes Add Pressure

Google’s AI division has undergone several leadership changes during a period of intense competition.

Demis Hassabis has stepped away from his previous day-to-day leadership responsibilities at Google DeepMind, while Koray Kavukcuoglu has taken on a greater role in daily Gemini-related decisions.

Google has also experienced departures involving prominent AI researchers, adding to the challenges facing the company as it tries to accelerate development.

Google Increasing Focus on AI Coding

Google has reportedly devoted more research resources and computing capacity to improving coding capabilities since the beginning of 2026.

The company has also strengthened its research team by hiring Barret Zoph, a former Thinking Machines Lab co-founder and OpenAI post-training lead. He joined Google as a vice-president of research focusing on reinforcement learning and post-training.

Google’s strategy involves working on multiple AI model approaches simultaneously, meaning setbacks involving one model do not necessarily indicate the performance of its broader AI pipeline.

Gemini Faces Growing Competition

Google is not competing only with OpenAI and Anthropic in AI coding. Several other technology companies are also developing tools aimed at software developers.

SpaceXAI, associated with Elon Musk, has introduced Grok models focused on coding. Meanwhile, Meta has launched Muse Code, an AI coding agent designed to assist with software engineering tasks.

The growing number of competitors means Google will need to demonstrate meaningful improvements in coding, reasoning and AI-agent capabilities.

Gemini’s Growing User Base

Despite the competitive pressure, Google has achieved significant adoption for its Gemini platform. The Gemini app has reportedly surpassed 1 billion users globally, demonstrating strong consumer interest in Google’s AI services.

However, the enterprise market remains particularly important. Businesses are increasingly adopting AI coding assistants and autonomous agents, creating an opportunity for Google to strengthen its position through more capable and cost-efficient models.

Conclusion

The reported launch of Gemini 3.8 Flash could become an important step in Google’s effort to regain momentum in AI coding. With internal tests reportedly showing promising results, the model could intensify competition with OpenAI and Anthropic while Google continues developing its larger Gemini models.