We benchmarked the top models on our own coding tasks.
The results: - GPT 5.6 Sol (high) won on performance - Grok 4.6 (high) was the runner-up - GLM 5.3 Flash won on cost at comparable quality All 50%+ cheaper than our previous default (Opus 5)
The AM Brief, Tuesday September 1, 2026 Part 3
Good morning, It’s 8am in Miami and here is a recap of events that caught my eye. Part 3 Anthropic published landmark research demonstrating that autonomous Claude agent teams can independently drive AI alignment,
My agents running Claude Fable 5.1 / Opus 5 / Qwen have been cooking HermesWorld Realms 🏰
A full-scale MMO 9 races, mounts, guild wars, forging, 58 zones, 20,227 textures, 7,845 animations. We're expanding the HermesWorldAI universe with a NEW game, with endless realms.
Fable 5.1 is cracked at game dev.
I’ve had it running in a loop for the last 24 hours trying to build a Fall Guys clone, and the progress so far is stunning. Fable 5.1 orchestrating Opus 5 worker agents with Sonnet 5 as the reviewer might be the best game dev stack I’ve tried
You have reached the end of the archive
All of Claude Opus 5