Google’s Gemini 4 Carbon ranks first on the Artificial Analysis Intelligence Index v5, scoring 59 and taking the top result in four of nine evaluations: HLE-Diamond, Terminal-Bench-Science, AA-Omniscience and AutomationBench-AA. Its strongest gains are in complex reasoning, scientific tasks, coding and knowledge work, though it does not lead every category. In coding-agent tests, Carbon scores 59 with Google’s Antigravity CLI, three points behind Claude Opus 5.5, and places second on both Terminal-Bench 4.0 and FrontierCode v1.1. Its AA-Briefcase score improves 126 Elo over Gemini 4 Argon, with particularly notable gains in presentation creation and analytical quality, but it remains behind several competing models.
At standard pricing, Carbon costs $7.76 per Intelligence Index task, about 23% less than Claude Opus 5.5, and uses 29% fewer output tokens per task; its cost is similar to Argon’s despite scoring nine points higher. Carbon also leads the updated AA-Omniscience evaluation, with 58% accuracy and an 11% hallucination rate - higher than Argon’s 4%, but far below the 50% rates reported for Claude Opus 5.5 and GPT-6 Astra. The model has a one-million-token context window and accepts text, images, video and speech. It is being offered to selected users, not yet publicly released, at $4 per million input tokens and $20 per million output tokens, with a 50% launch discount and 95% off cached inputs.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
GitHub Copilot Fusion routes coding tasks among Anthropic and OpenAI models to balance performance, cost, and speed. Benchmarks found it scored two points below Claude Opus 5.5 while costing $5.59 per task - about half as much - and using roughly half the tokens. (twitter.com)
Doomformer compiles an integer neural network into a DOOM map, using voodoo dolls for arithmetic and lighting and floor heights to store values. It checks every result against NumPy with zero tolerance. (github.com)
Dartmouth’s provost resigned following accusations that he used AI for writing. The case raised questions about the consequences of using AI in professional work. (apnews.com)
Z.ai’s open-weight GLM 5.3 model exploited vulnerabilities in six recent Firefox releases using security fixes published before users received them. Each attack cost about $1,000, showing how AI can turn the patch gap into a practical security risk. (epoch.ai)
Alphabet lost just over $500 billion in market value after investors reacted to news that AI summaries are affecting Google Search ad revenue. Heavy AI spending by Google, Microsoft and Amazon also raised concerns about growth in their core businesses. (bbc.com)
Nanbeige4.3-33B is a 33-billion-parameter flagship language model developed by BOSS Zhipin. The provided page content includes its chat template but no performance details. (huggingface.co)
Today's best Hacker News stories, summarized and screenshotted, one email a day.