Sooooo Mythos is basically gpt 5.5 but it was so dangerous it could create an apocalypse with nuclear weapons and rip through computer systems as we know it today 🤔

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…

After two days with Claude Fable 5 the best way I can describe it is "relentlessly proactive" - here's an example where I dropped in a screenshot of a bug and it span up custom CORS Python servers and used pyobjc-framework-Quartz to capture screenshots simonwillison.net/2026/Jun/1…

I can't help but feel personally burned by the Claude Code changes announced today. We put so much work into wrapping the (atrocious) Claude Agent SDK in T3 Code. It was the ONLY path they supported, so we made it work. It was hell. Now our users are getting their rate limits cut by 40x, despite us doing everything right. I listened to the Claude Code team. I had my issues with their direction, but I trusted them and took them at their word. I will never make that mistake again. Until we see significant change, it is safe to assume any statement from an Anthropic employee is a lie on a timer. The rug will be pulled, no matter how many promises are made beforehand.

I ran Qwen3.6:27b to optimize itself in a recursive loop on my home server. Over 26 hours it went from 2.3tok/s to 84.3tok/s decode. It began on the home server, found there was no NVIDIA GPU, detected a CPU/RAM setup with 24 CPU threads and 93 GiB RAM, a 9060xt 16gb, then installed Hugging Face tooling remotely and started pulling GGUF quantizations. It benchmarked remote llama.cpp / llama-server runs across quantizations and flags: Found existing Qwen3.5-9B-Q8_0.gguf Downloaded / tested Qwen3.6-27B GGUF variants Compared Q6_K, Q5_K_M, Q4_K_M Ran server benchmarks over SSH against localhost:8080 Tested thread count, context, batch size, n_ubatch, --no-mmap, and memory-related flags Researched further speed paths: lower quantization, NUMA, huge pages, native CPU builds, cache/KV, TurboQuant, DFlash, speculative decoding, and automated tuning 1,524 tool calls 367 artifacts 345 memory addition 804 browser-control calls All of this from a model that can run on your computer. You don't need a better model. You need a better harness.