Fetching from the wire…
01
02
03
04
05
06
07
08
Models2026-08-13 · source-backed
Simon Willison found it in the OpenRouter price list. 1.6T MoE, ~49B active, 1M context, up to 384K output, at $0.435/M input (cache miss) and $0.87/M output. The agentic-coding deltas versus the preview are the story: DeepSWE 12.8 → 62.7, CyberGym 52.7 → 83.3, Terminal Bench 2.1 72.1 → 87.9. Near-frontier agent coding at roughly 1/60th of frontier pricing, with a significant API price increase already announced. Willison also notes the three reasoning-effort levels produce visibly different SVG output on his pelican benchmark, which he hasn't seen from other configurable-effort models. Meanwhile the Hugging Face repos still host April preview artifacts and the API endpoint swapped underneath the same model ID: Vercel's AI Gateway picked up new weights by default, so existing code changed silently.
Each link below shares sources, entities, or timing with this story.
DeepSeek released V4 Preview on April 24 with two open-weight variants: V4-Pro (1.6T total parameters, 49B activated via MoE) and V4-Flash (284B parameters, 13B activated). Both support 1M-token context windows. Both are Apache 2.0 licensed. Both are live right now on Hugging...
Satya Nadella said companies routing everything through a single proprietary lab may not survive. His argument: you hand that lab your most sensitive business context, and the lab can turn it against you as a competitor. His prescription is an orchestration layer — keep the ha...
GPT-5.6 Luna went to $0.20 input / $1.20 output per million tokens on July 30. That's an 80% cut. Terra dropped 20%. Luna's input now undercuts Gemini 3.1 Flash-Lite ($0.25/$1.50) and sits at one-fifth of Claude Haiku 4.5's $1 input. Simon Willison covered the announcement and...
Simon Willison doesn't hand out superlatives. So when he writes that Z.ai's GLM-5.2 is "probably the most powerful text-only open weights LLM," that's worth stopping for. His June 17 evaluation walks through a 753B-parameter Mixture-of-Experts model with 40B active params, a 1...
Alibaba released Qwen3.6-27B on April 22. Dense architecture. Open weights. 77.2% on SWE-bench Verified, within 3.7 points of Claude Opus 4.6. On SkillsBench, it scores 48.2% versus its own 397B MoE predecessor's 30.0%. That's a 77% improvement with 14.8x fewer parameters. Let...
$1.25 input / $4.25 output per million tokens for the standard tier. $0.10 / $0.20 for the Contributor tier, where Meta trains on your usage and feedback. That's roughly 12x on input, 21x on output, and it's the clearest number anyone has published on what your proprietary sou...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.