AI benchmarks

AI benchmarks, newest first

1. 

   AI

   Meta Launches Muse Wonder and Wonder Full Frontier AI Models

   Meta introduces Muse Wonder and Wonder Full, frontier AI models with active memory, advanced coding and access through Muse Code and its developer API.

   By Shivangi Yadav12 October 20263 min read
2. 

   Gaming

   GPT-6 Astra Downloaded a Human-Made StarCraft Bot After Its Own Code Fell Short

   GPT-6 Astra downloaded the human-built Stardust bot during a StarSkirmish test after struggling with stronger opponents, exposing a challenge for AI agents.

   By Saumya Yadav5 October 20267 min read
3. 

   AI

   Claude Fable 5.1 Is Out, And You Must Read About It

   Claude Fable 5.1 explained, covering AI benchmarks, biological research, safety findings, hidden-task tests, watermarks, costs, and new capabilities.

   By Nihal Singh3 September 20268 min read
4. 

   AI

   Anthropic Releases Fable 5.1 With Lower Costs and Fewer False Restrictions

   Anthropic launches Fable 5.1 with lower token costs, fewer false-positive safeguards, privacy updates, and improved AI benchmark performance.

   By Shivangi Yadav2 September 20263 min read

5. 

   AI

   GPT 6 Astra Might Dethrone Opus 5.1

   GPT 6 Astra could challenge Opus 5.1 with stronger reasoning, coding and agentic capabilities, potentially reshaping the race among the leading AI models.

   By Nihal Singh31 August 20267 min read
6. 

   AI

   How DeepSeek V4 Pro became a serious rival to the world's best AI models

   DeepSeek V4 Pro explained, from its efficient architecture and pricing to benchmarks, open models, and why it is changing how AI systems compete fast.

   By Nihal Singh27 August 20267 min read
7. 

   AI

   Nvidia’s AVO Harness Pushes Claude Opus 5 to 100% on ARC-AGI-3

   Nvidia’s AVO harness pushed Claude Opus 5 from a 30% baseline to a 100% ARC-AGI-3 score, highlighting the role of agent architecture in long AI tasks.

   By Shivangi Yadav22 August 20264 min read
8. 

   AI

   Arena AI Leaderboard Reaches $100 Million in Annual Revenue After Rapid Growth

   Arena, the widely used AI leaderboard platform, has reached $100 million in annualised revenue just months after launching its commercial AI evaluation service. Learn how Arena became a…

   By Shivangi Yadav1 July 20263 min read

9. 

   Brief

   Google’s new Gemini Pro model has record benchmark scores — again

   Google’s latest Gemini Pro model has posted record benchmark scores across coding, reasoning, and multimodal tests, intensifying competition in advanced AI systems.

   By Shivangi Yadav21 February 20262 min read
10. 

    AI

    China’s Moonshot Releases Open-Source Multimodal Model Kimi K2.5

    Moonshot AI has launched Kimi K2.5, an open-source multimodal AI model that handles text, images, video, and coding tasks, with performance competitive with leading proprietary systems.

    By TechAmerica.ai Staff27 January 20263 min read
11. 

    Startups

    LMArena lands $1.7B valuation four months after launching its product

    LMArena has raised $150 million in a Series A round at a $1.7 billion valuation, just four months after launching its commercial AI evaluation platform.

    By TechAmerica.ai Staff6 January 20263 min read
12. 

    AI

    OpenAI Upgrades Codex with a New Version of GPT-5

    OpenAI releases GPT-5-Codex, a new AI coding model for Codex, enhancing dynamic task processing and outperforming previous benchmarks. Available for ChatGPT Plus, Pro, Business, Edu, and…

    By TechAmerica.ai Staff15 September 20252 min read