Google launched Gemini 3.8 Flash in September 2026, marking AI momentum after a prolonged stock decline, with the model emphasizing coding and cybersecurity while competing on pricing against rivals like Anthropic and OpenAI.
Alphabet’s stock edged higher in early September 2026 after a prolonged losing streak, buoyed by the launch of Gemini 3.8 Flash, Google’s latest AI model focused on coding and cybersecurity. The company also secured a legal victory as a federal judge rejected the Justice Department’s bid to force Google to sell its ad exchange, easing antitrust pressures. However, the stock remained slightly down for the month, reflecting ongoing challenges in the AI race.
Gemini 3.8 Flash: A Push for Enterprise Revenue
Google positioned Gemini 3.8 Flash as its most advanced reasoning and coding model yet, with improvements over its predecessor, 3.7 Flash, in software engineering and multi-step tasks. Tulsee Doshi, senior director of product management at Google DeepMind, told CNBC the model really surprised us in positive ways in their performance,
adding that it gives us opportunities to lean into them.
The model is designed to appeal to enterprise customers, with Google emphasizing its cost-effectiveness compared to larger rivals.
At 75 cents per million input tokens and $3.75 per million output tokens, Gemini 3.8 Flash matches the introductory pricing of its predecessor, even as it offers enhanced performance. Google also introduced pay-as-you-go pricing for Gemini Enterprise, along with discounts and spending caps, aiming to undercut competitors like Microsoft and Anthropic, which rely on recurring seat fees and separate product licenses. We’re really excited about being able to provide an offering to defenders that is a fraction of the cost, much faster, while still showcasing that frontier-level performance,
said Doshi.
Analyst Reactions: Momentum, But Not a Market Shift
Google’s scale remains a key asset. Nearly three-quarters of Google Cloud customers already use its AI products, with CEO Thomas Kurian reporting that these users spend about 50% more than their original commitments. DeepMind’s Demis Hassabis also outlined a vision for Gemini as a general-purpose layer coordinating cheaper, specialized models, suggesting breadth could offset the need for the “best” single model.
Stock Performance and Legal Wins
Alphabet’s stock rose 0.6% on the day of the Gemini 3.8 Flash announcement but remained slightly lower for September after dropping over 1% at the start of the month. The company’s broader AI strategy gained traction with a federal judge’s rejection of the Justice Department’s antitrust case, a setback for regulators seeking to break up Google’s ad exchange. Berkshire Hathaway CEO Greg Abel also praised Alphabet’s AI position, citing increased interest from portfolio companies that rely on Google’s technology.
Despite these gains, Google faces headwinds. The company’s AI investments have driven massive capital expenditures, with spending counting on growth and market share gains to deliver returns over the longer term. We have a lot of visibility from within our companies as to how we’re using AI, what type of benefits it’s delivering,
Abel said, signaling confidence in Alphabet’s long-term trajectory.
Google’s cybersecurity model, Gemini 3.8 Flash Cyber, aims to differentiate the company further by offering frontier-level performance
at a lower cost. However, access is limited to a small group of trusted government and enterprise defenders
through its Fairwind Program, reflecting concerns about potential misuse.
The AI race remains fiercely competitive, with Google’s latest moves highlighting its strategy of balancing performance, pricing, and scale. While Gemini 3.8 Flash represents progress, the company’s ability to translate this into market share and profitability will depend on its execution and the broader AI landscape’s evolution.
