Gemini 3.5

Evidence (13)

• 7Ars Technica AI·10/1/2026regulation
  Judge dismisses Chegg and Penske antitrust cases over Google AI search

  A US federal judge dismissed antitrust lawsuits filed by Chegg and Penske Media against Google over AI search features such as AI Overviews. The ruling found that Google's conduct was not illegal under antitrust law, despite claims that AI-generated answers reduced referral traffic to publishers and content sites.

• 8Ars Technica AI·9/30/2026model_release
  Google unveils Gemini 4 Argon, not yet available to users

  Google has announced Gemini 4 Argon, a new frontier AI model that it says leads on coding, knowledge work, and cybersecurity benchmarks. The company says it is already being used internally to save memory across data centers and to help migrate large C/C++ codebases to Rust, but external access is not yet available.

• 7Hacker News·9/30/2026model_release
  Gemini 4 Argon Discussion

  This Hacker News item points readers to a separate analysis of Gemini 4 Argon covering its intelligence, performance, and pricing. The post generated substantial discussion, with high points and many comments, suggesting strong community interest in the model and its trade-offs.

• 7Hacker News·9/24/2026security
  Hackers exploit ChatGPT and Gemini to steer users to scam centers

  A Hacker News-linked article reports that attackers are manipulating AI chatbots such as ChatGPT and Gemini so they provide directions or guidance that can lead users to scam centers. The piece highlights a prompt-injection style abuse of generative AI systems rather than a direct compromise of the underlying models.

• 2Simon Willison·9/23/2026research
  Gemini 3.8 adds text-to-speech greeting

  Google DeepMind says Gemini 3.8 can generate text-to-speech output that says hello. The announcement indicates a new speech capability in the Gemini 3.8 model, but provides no further technical details in the headline alone.

• 5Ars Technica AI·9/17/2026product_launch
  Google unveils experimental CC AI agent for families

  Google Labs has introduced CC, an experimental AI agent designed to be shared by up to six members of a family. The agent uses a dedicated Google account and can interact with shared or explicitly granted content from Gmail, Google Chat, and a shared Google Drive folder.

• 4Simon Willison·9/15/2026model_release
  Google DeepMind introduces Gemini 3.8 Live and Extended Thinking

  Google DeepMind announced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. The update appears to add a live mode alongside a longer-reasoning variant, but the provided announcement text contains no further technical details.

• 7The New Stack·9/14/2026research
  Real-SWE benchmark shows top coding agents fail most of the time

  The Real-SWE benchmark tested coding agents on private, real-world company codebases and found that Claude Fable 5.1 led with only 38.8% success, still failing more than 60% of attempts. Results varied sharply by model and task, and no agent was consistently reliable across unfamiliar production code.

• 6Hacker News·9/3/2026research
  Study compares Claude, Codex, and Cursor tool choices across 17k runs

  The article describes a large-scale experiment measuring which tools coding agents choose across thousands of realistic repository tasks. The researchers built a panel of 75 repositories, ran 16,893 sessions in ephemeral sandboxes, and used simulated-human and judge agents to evaluate valid runs and final winners.

• 8Hacker News·9/2/2026model_release
  Google Releases Gemini 3.8 Flash and Cyber Variant

  Google DeepMind has published model cards for Gemini 3.8 Flash and a cybersecurity-focused Gemini 3.8 Flash Cyber variant. The HN discussion centers on the model release, its capabilities, and how the cyber-tuned version may be used for security-related tasks.

• 6Hacker News·8/13/2026research
  How AI Text Watermarking Works

  The article explains that AI text watermarking does not hide marks in characters or metadata, but in the model’s word-choice decisions. It describes how secret keys bias word selection so generated text can later be detected by counting patterns in the same choices, even after copying.

• 5Simon Willison·8/13/2026framework_update
  llm-gemini 0.33 adds Gemini 3.7 Flash support

  The llm-gemini plugin version 0.33 adds support for Gemini 3.7 Flash, along with Gemini 3.6 Flash, Gemini 3.5 Flash Lite, and two embedding models. It is also updated for LLM 0.32 compatibility, enabling reasoning traces and server-side tools such as CodeExecution.

• 6Hacker News·8/13/2026model_release
  Google rolls out Gemini 3.7 Flash

  Google has begun rolling out Gemini 3.7 Flash, replacing Gemini 3.6 Flash only three weeks after its release. The company says the new model is optimized for coding, agentic tasks, and document processing, and is being launched with a lower introductory price to stay competitive.