Google releases updated Gemini 2.5 Pro, says it’s the “most intelligent model yet”

Google is releasing an updated Gemini 2.5 Pro model, addressing performance criticisms from previous updates, particularly outside of coding tasks. The new version, expected to become a stable release for the Gemini app, boasts improved coding capabilities, achieving an 82.2% score on the Aider Polyglot test, surpassing competitors like OpenAI and Anthropic. This release also includes configurable thinking budgets for developers and aims to restore the general performance levels seen in the 03-25 update.

Analysis

Google is rolling out a significant update to its Gemini 2.5 Pro AI model, primarily focused on addressing performance regressions observed in non-coding tasks since the 03-25 update and enhancing its coding capabilities. This new version, designated 06-05, is slated to become a stable, long-term release, eventually shedding its "Preview" status for the consumer-facing Gemini app and web interface. Notably, Google reports a new high score of 82.2 percent in the Aider Polyglot coding test for this iteration, surpassing leading competitors such as OpenAI, Anthropic, and DeepSeek. According to Google's Logan Kilpatrick, the update specifically targets user feedback regarding performance degradation, promising improvements in creativity and response formatting. Furthermore, the 06-05 release introduces configurable thinking budgets for developers, indicating a strategic move to improve developer experience and model efficiency. This update signals Google's commitment to refining its flagship AI model's general-purpose utility while maintaining a competitive edge in specialized areas like code generation.

Key Decisions for Investors

• Monitor the upcoming stable release of Gemini 2.5 Pro to assess its real-world performance improvements and user adoption, as this could positively impact Alphabet's (GOOGL, GOOG) competitive positioning in the AI sector.
• Consider the enhanced coding capabilities and the claim of outperforming competitors as a potential driver for increased enterprise adoption and developer preference, which may translate to future revenue streams.
• Evaluate the impact of this update on the broader AI market, particularly how it affects the competitive landscape with OpenAI and Anthropic, as Google's progress could influence investment sentiment across the AI theme.