Would be interesting to eval how opus 5 performs on tasks autonomously vs iterative tasks
GLM 5.3, GPT 5.6 Luna, Claude Opus 5, Grok 4.6, DeepSeek V4, Kimi K3, and a cloud browser for agents - all FREE.
DuckDuckGo: GPT 5.6 Luna and older models, no signup: LM Arena: free side-by-side, Opus 5, GPT-5.6 Sol, Grok 4.6, Qwen 3.8 Max:
Claude Opus 5 built this school of manta rays in p5.js
It’s a single HTML file with no textures or premade models. All the geometry is recalculated from scratch every frame The wings are built along their own arc length, so they don’t stretch while flapping. A traveling wave
Claude fable 5.1 is here
Terminal-Bench 4.0: Fable 5.1 → 55.8% Opus 5 → 52.3% Fable 5 → 42.0% Cache reads: $1.00 → $0.25/MTok Anthropic says typical workloads cost ~25% less, and highly agentic ones up to ~45% less Agents just got a serious upgrade
You have reached the end of the archive
All of Claude Opus 5