Claude Opus 4.7

Epoch introduced MirrorCode, a benchmark where models reimplement real programs from specs with no internet and hidden held-out tests; the best current score is 56%. The setup matters because it scales inference into multi-day runs and targets software jobs estimated to take humans weeks.

Genspark turned Build Preview into Genspark Design and merged its AI Designer tooling into one product with Figma uploads, reusable brand systems, and code export. The launch matters because it pushes design-to-code workflows toward editable layered output instead of one-shot mockups.

Arena shipped Agent Mode, a benchmark that lets models use web search, bash, file writing, image generation, and follow-up questions, then ranks them on five live-session signals. It matters because agent evals move from static task sets to real user workflows, with GPT-5.5 High currently leading the leaderboard.

Claude Code 2.1.154 added Dynamic Workflows, a research-preview mode that writes orchestration scripts and runs hundreds of subagents in one session. Anthropic also shipped 2.1.156 to fix Opus 4.8 thinking-block API errors, so teams should watch for workflow and API stability.

Anthropic released Claude Opus 4.8 across Claude, the API, and major clouds with higher coding scores and a cheaper 2.5x-speed Fast mode. Use it for coding workloads that want better benchmark performance without a price increase over 4.7.

DeepSWE launched a coding benchmark built from 113 original tasks across 91 repos and five languages, with GPT-5.5 leading at 70%. The setup is meant to better reflect repo search, multi-file edits, and verification in real agent workflows.

Cline open-sourced the runtime behind its extension and CLI as the Cline SDK, then rebuilt the CLI on top with agent teams, cron jobs, connectors, and example apps. The harness score gives teams a new reference point if they want to compare agent tooling on Terminal-Bench 2.0.

Anthropic rolled fast mode for Opus 4.7 into Claude Code and tools including Cursor, v0, Droid, Conductor, and OpenRouter. Use it where latency matters, but watch pricing: Cursor disclosed a 6x multiplier and others treat it as premium.

User posts and HN threads compared GPT-5.5 and Opus 4.7 across plan mode, frontend work, and 120K-context sessions. The split results mean token burn and instruction discipline matter as much as raw benchmark scores.

Users on Hacker News and Reddit reported a reproduced HERMES.md extra-usage billing bug, plus new ban appeals and repeated blame-shifting complaints. Anthropic says affected users will get refunds and credits, so teams should keep an eye on quota routing and support escalation.

Days after Opus 4.7 launched, users reported commit-message triggers tied to OpenClaw or HERMES markers that could route requests into extra billing or refusals, alongside continued throttling complaints. Anthropic says affected users will get refunds, but repo-scanning heuristics may still affect cost and reliability in multi-harness workflows.

ARC Prize published frontier-model results on ARC-AGI-3 and said GPT-5.5 and Opus 4.7 both stayed below 1%, with failures in world modeling, abstraction, and reward reinforcement. That shows strong coding and benchmark models still break on novel interactive reasoning tasks, and follow-up comparisons even had Opus 4.6 slightly ahead of 4.7.

Anthropic opened Claude Security to Claude Enterprise customers, letting teams scan repositories, validate findings, and review suggested patches inside Claude. The beta also adds scheduled scans, directory targeting, exports, and webhook alerts for recurring codebase reviews.

A day after Opus 4.7 launched, users reported OpenClaw-linked refusals, cache TTL cost spikes, and billing failures in Claude Code. Anthropic appears to have eased some limits, but behavior and spend still vary sharply across agent-heavy sessions.

Users reported more verbosity, weaker 1M-context behavior, and little coding gain after Opus 4.7 rolled out. OpenRouter measured 12–27% higher costs, and some teams reverted their default model.

Anthropic said three harness-side changes degraded Claude Code quality, then reset subscriber limits and rolled out fixes in 2.1.119. The update matters because recent failures came from tool defaults and prompt handling rather than the base model alone.

Four days after the Opus 4.7 launch, independent tests measured about 1.35-1.46x more text tokens than 4.6 while users kept reporting faster limit burn and weaker coding. That can change effective cost and session economics in Claude Code even if list prices stay flat.

Users and analysts say Opus 4.7 is using more tokens, refusing web search, and missing orchestration steps in Claude Code-style workflows. Watch token costs and regression reports closely if you rely on xhigh defaults or tokenizer-sensitive prompts.

Anthropic launched Claude Design in research preview, turning prompts, files, and codebase context into prototypes, slides, and one-pagers. It can infer a team design system and export to Canva, PDF, or PPTX, or hand off to Claude Code.

A day after Opus 4.7 launched, users are surfacing adaptive-thinking misses, surprise refusals, and higher token use. For engineers, recheck prompts, costs, and 4.6 fallbacks while Anthropic patches bugs and lifts limits.