Grafy News

Google releases Gemini 4 Argon, challenges AI frontier

Google has launched its long-awaited Gemini 4 Argon model, claiming it competes with the industry's best in reasoning and complex workflows.

Google Launches Gemini 4 Argon

After months of trailing competitors in the artificial intelligence sector, Google has officially released Gemini 4 Argon. The company claims this new model can compete with the best offerings in the industry. According to Robert Hart, an AI reporter at The Verge, Argon delivers "frontier reasoning and improved performance across complex workflows." This launch marks a significant step for Google as it attempts to reclaim its position against rivals like OpenAI and Anthropic.

Performance Claims and Benchmarks

Google states that Gemini 4 Argon outperforms rival models such as GPT-6 Astra, Claude Fable 5.1, and Claude Opus 5.5. These superior performances are noted across a variety of benchmarks including coding, cybersecurity, science, finance, and law. The model aims to reduce common issues found in previous AI chatbots, such as hedging, repetition, and clichés. By improving these areas, Google hopes to demonstrate tangible advancements in natural language processing and task execution capabilities.

Internal Use Cases

The technology is already being utilized internally by Google. Specific applications include finding and patching cybersecurity vulnerabilities within their systems. Additionally, Argon has been deployed to halve the computing power needed to manage YouTube’s comments section, optimizing resource usage significantly. Another internal use case involves debugging the drone physics simulator for Wing, Alphabet’s drone delivery company. These examples illustrate practical implementations of the model in real-world scenarios beyond just theoretical benchmarks.

Pricing Strategy

Google is introducing Argon with a highly competitive pricing structure. The introductory rate is set at $2 per million input tokens and $10 per million output tokens. This cost is approximately one-fifth of what OpenAI’s GPT-6 Astra and Anthropic’s Fable 5 models charge. Although these prices are scheduled to double at the end of November when the introductory period concludes, they will still remain significantly cheaper than frontier rivals on a per-token basis. However, it is important to note that different models may use varying numbers of tokens for identical tasks, so actual cost savings might differ in practice.

Availability and Access

The rollout of Gemini 4 Argon began today for specific user groups. It is available to Google’s higher-tier Pro and Ultra subscribers through the Gemini app and AI mode in Search. Furthermore, Ultra subscribers gain access via Antigravity, Google’s agentic development platform. There is currently no clarity regarding availability for Plus subscribers or free users. This tiered approach suggests a strategy to reward premium users while gathering extensive data from high-volume professional environments before wider dissemination.

Context Within the AI Race

This release follows a period where Google trailed behind competitors due to impressive models from Anthropic and OpenAI. Seismic internal shakeups had raised doubts about Google’s ability to compete at the frontier. However, Argon’s arrival indicates that Google remains a serious contender. Notably, this launch comes just a day after Meta announced its own frontier model, Muse Wonder Full. This timing highlights how quickly the balance of power can shift in the AI landscape, proving that the field is not merely a two-horse race between established giants but includes dynamic players constantly pushing boundaries.

Conclusion

Google’s release of Gemini 4 Argon signifies a strategic move to reassert dominance in the AI arena. With claimed improvements in reasoning, workflow performance, and cost efficiency, the model presents a compelling option for developers and enterprises. As the competitive landscape continues to evolve with simultaneous announcements from other tech leaders, the impact of Argon on market dynamics remains to be seen. Users interested in testing these capabilities can explore similar functionalities using tools like Grafy Chat.


Source: The Verge — AI — Google finally releases Gemini 4 Argon