Tech News
← Home  ·  All topics

Opus 5.5

59 GoKawiil briefs on this topic

Google's Gemini 4 Carbon edges Claude Opus 5.5 on new intelligence benchmark

Google has launched Gemini 4 Carbon, which Artificial Analysis now ranks as the top AI model on its Intelligence Index V5 suite, narrowly surpassing Anthropic's Claude Opus 5.5 by one point and besting OpenAI's GPT-6 Astra and Meta's Muse Wonder Full. The model also leads a separate Artificial Analysis test measuring hallucination rates and factual recall, though its score of 54 out of a possible 100 shows even the top model remains far from fully reliable. Google is initially rolling the model out to cyber defenders via its Fairwind program, with broader access to follow later.

Anthropic adds Claude Code prompt barring recreations of specific existing video games

Anthropic quietly inserted a new system-prompt rule into Claude Code, first spotted around October 27 and confirmed present in version 2.1.314 released October 30, instructing the model not to rebuild a specific existing game even if renamed. Anthropic engineer Addy Osmani confirmed the company added 'two sentences' to steer Claude Opus 5.5 away from recreating someone else's game, citing its usage policy against IP infringement, while still permitting original or genre-based games built from a user's own assets.

Anthropic adds prompt restrictions to Claude Code against cloning specific games

Anthropic revised the system prompt for Claude Code on its Opus 5.5 model, inserting instructions directing the AI to avoid recreating specific existing video games. The company says the change enforces its usage policy against using Claude to infringe intellectual property, while original games, genre-style games, and projects built from a user's own designs and assets should still function normally.

Anthropic confirms Claude Code was updated to block cloning of existing video games

Anthropic developer advocate Addy Osmani confirmed the company added two sentences to Claude Code's Opus 5.5 prompt instructing it to avoid recreating a specific existing game, even if renamed, citing its usage policy against IP infringement. He said original games, genre games, and projects built from a user's own designs and assets should still work, and that Anthropic will keep adjusting the rules. The change follows developer complaints on GitHub that Claude Code was refusing game-building and reverse-engineering tasks.

Developer uses Claude AI to clone Adobe apps as free Artcraft suite

Developer Brandon Thomas built Artcraft, a set of free, open-source clones of Adobe Creative Cloud apps including a Photoshop substitute called Photocraft, using Claude Opus 5.5 to reverse-engineer the software in Rust. Thomas says the move was prompted by frustration with Adobe's subscription cancellation fees, and the apps mimic Adobe's interfaces closely enough to require little relearning. Testing found Photocraft handles basic tasks like resizing and masking well but suffers from bugs, instability, and an inability to open native PSD files.

Developer builds HackerTube, a YouTube alternative sourced from Hacker News comments

A developer used Anthropic's Claude Opus 5.5 model to build HackerTube.org, a YouTube-style site populated only with videos that commenters recommended on Hacker News. The project scraped HN comments from 2007-2026, filtered roughly 40,000 video recommendations down to 6,454 videos across 150 topics using Claude Sonnet, then wrapped the results in a YouTube-like interface with custom recommendations stored locally rather than via user accounts.

Claude models block game recreation and reverse-engineering tasks, users report

Users say Anthropic's Claude Code and Opus 5.5 are refusing tasks tied to reverse-engineering, including use of the Reverse Engineer Anything tool, apparently due to a new internal rule against recreating existing games. A purported rule text circulated by a Twitter user states the model should build original games but not recreate a specific existing game even if renamed. Developers have also reported refusals mid-project, when replicating mechanics, or patching tabletop-style games, alongside separate reports of refusals on unrelated reverse-engineering tasks like analyzing crash dumps or removing TV junkware.

Anthropic's Duet assistant draws praise for speed and channel-based organization

A reviewer describes hands-on use of Duet, an Anthropic-linked personal-assistant chat tool reportedly built on the Opus 5.5 model, after testing several competing assistant apps like Muse, Dot and NanoClaw. The reviewer highlights three features: faster response times than other Claude interfaces, automatic creation and management of conversation 'channels' to organize work, and per-channel access restrictions to GitHub repositories that avoid requesting user credentials.

Google's Gemini 4 Carbon takes #1 spot on Artificial Analysis Intelligence Index v5

Artificial Analysis reports that Google's new Gemini 4 Carbon model has topped its Intelligence Index v5 with a score of 59, ranking first in 4 of 9 component evaluations including HLE-Diamond and AA-Omniscience. The model also placed second on the firm's Coding Agent Index, trailing only Anthropic's Claude Opus 5.5, and showed sizable gains in agentic knowledge-work benchmarks over its predecessor, Gemini 4 Argon. The model is currently in limited rollout with a 50% launch discount and not yet publicly available.

Developer builds Pokemon-style Game Boy Color interface as custom TV home screen

A developer created a custom TV home screen designed to resemble a Pokemon town rendered in Game Boy Color style, using in-game actions as navigation. Entering buildings in the interface opens apps, and changing device settings requires defeating a gym leader within the mock game world. The project was reportedly built with the help of Anthropic's Opus 5.5 AI model.

Moonshot AI launches Kimi K3.1 Pro with 1-million-token context window

Moonshot AI released Kimi K3.1 Pro, an open-weight 3T-class model built on Kimi K3 with expanded post-training compute, offering a 1-million-token context window and gains in coding, 3D creation, financial research and agentic tasks. A faster variant, K3.1 Pro Highspeed, runs 2-3 times quicker and both are live now on Kimi.ai, Kimi Work, Kimi Code and the Kimi API, with full model weights due by November 4.

Anthropic launches Claude Duet, an agentic assistant app for professionals

Anthropic has released Claude Duet, a new app built on Opus 5.5 that connects to work tools like Gmail, calendar and Slack to proactively manage tasks for professionals. The app flags priorities, surfaces what needs attention, and indicates what can wait, while including safeguards that block or confirm before sharing sensitive data. Duet is currently in research preview, limited to Pro and Max subscribers, with a waitlist open for others.