Gemini 3

1. My dream of a fully generative game is getting pretty close to possible now. I made a demo where you can prompt any spell and fight online. (www.reddit.com)

   +5718 7w gemini

   Prompt any spell and use it in a 3D physics based world, powered by Gemini 3 Full multiplayer support for up to 6 players with VoIP All made with ThreeJS and Colyseus https://spellwright.xyz/

2. ChatGPT/Gemini can now draw on your screen to help you navigate complex software (sketchvlm.github.io via hn)

   +61 8w gpt-5geminichatgpt

   When answering questions about images, humans naturally point, label, and draw to explain their reasoning. In contrast, modern vision–language models (VLMs) such as Gemini-3-Pro and GPT-5 typically respond with only text, which can be diff…

3. Show HN: Prediction market analysis app layering LLMs with data APIs (apps.apple.com via hn)

   +22 8w sonnetgemini

   I created a prediction market analysis app after trying prediction markets and doing quite poorly. I wondered if AI-driven predictions could be better with the right data.

4. Show HN: OSS Agent I built topped the TerminalBench on Gemini-3-flash-preview (github.com via hn)

   +15 8w gemini

   Scored 65.2% vs google's official 47.8%, and the existing top closed source model Junie CLI's 64.3%. Since there are a lot of reports of deliberate cheating on TerminalBench 2.0 lately (https://debugml.github.io/cheating-agents/), I would…

5. Anyone else noticing how Gemini-3-Flash is becoming the 'hidden' beast for automated promotions, its so productive? (www.reddit.com)

   +32 9w geminiagentic

   I've been testing a few different models for desktop-driven outreach and promotion workflows. While everyone is eyeing the massive LLMs, Flash-Preview is hitting that sweet spot of speed and reliability for multi-step agentic tasks and its…

6. GPT-Image-2 vs Gemini-3-Pro-Image (www.reddit.com)

   20 9w geminiopenai

   I feel like OpenAI still have some work to do. What do you think?

← all threads