Simon Willison on llm-pricing
73 posts tagged ‘llm-pricing’. Posts about the pricing of various LLMs. See also my pricing calculator.
Introducing Claude Sonnet 4.5
anthropic.comClaude Sonnet 4.5 is the best coding model in the world, strongest model for building complex agents, and best model at using computers.
Claude Sonnet 4.5 is probably the “best coding model in the world” (at least for now)
Anthropic released Claude Sonnet 4.5 today, with a very bold set of claims: Claude Sonnet 4.5 is the best coding model in the world. It’s the strongest model for building …
Anthropic’s Claude Sonnet 4.5 Can Code Autonomously for 30+ Hours
Anthropic may be using its specialization in coding and STEM as a differentiator in the buzzy frontier model market.
Anthropic just changed the rules for working with AI, and "prompting" isn't the main game anymore.
Anthropic published a new guide on "context engineering" that fundamentally shifts how we should think about getting better results from AI. Turns out, the words in your prompt matter less than what information you're feeding the AI in the first place.
Alignment is not solved
But it increasingly looks solvable
Import AI 431: Technological Optimism and Appropriate Fear
What do we do if AI progress keeps happening?
Import AI 431: Technological Optimism and Appropriate Fear
Welcome to Import AI, a newsletter about AI research. Import AI runs on lattes, ramen, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Import A-IdeaAn occa…
The AI Gründerzeit
Notes from Three Weeks in the Valley
Speeding into the AI of the hurricane
Substack: We’re speeding directly into the AI of the hurricane In three words, yesterday was wild.
The Secret Prompts in GitHub Copilot CLI - Will Vincent
Sleuthing in the source code of GitHub's new AI CLI.
Claude Sonnet 4.5 knows when it’s being tested
Anthropic's new model appears to use "eval awareness" to be on its best behavior
Import AI 431: Technological Optimism and Appropriate Fear
Welcome to Import AI, a newsletter about AI research. Import AI runs on lattes, ramen, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now Import A-IdeaAn occa…
Managing context on the Claude Developer Platform | Claude
New context editing and memory tools enable Claude agents to handle long-running tasks without hitting context limits.
I Vibe Coded a WordPress Plugin and Shipped it to Production - Kerrick Long (blog)
I got a buggy WordPress plugin from ChatGPT in 15 minutes. Then I got a good plugin in one shot from Claude. See my failure, success, and iterative improvement.
AI coding is now everywhere. But not everyone is convinced.
Developers are navigating confusing gaps between expectation and reality. So are the rest of us.
Big Players in AI - dentro.de/ai
The Who is Who in AI Land.
Claude Skills in Claude Code: A Compleat Guide
I migrated my Cursor Rules to Skills in Claude Code. Simple, elegant, powerful.
When Claude Code Skates on Thin Ice
I love Claude Code and Sonnet 4.5, but when training data is thin, the reasoning failures are hilarious
Your Coding Agent Is a Slot Machine
Measuring run-to-run variance in coding agents: Claude Code vs Mistral's Vibe.
State of LLMs in Late 2025
Analyzing the current LLM landscape and its use-cases
Councils of agents: group thinking with LLMs - The Engineering Manager
Improving your thinking is improving your output as a leader. Use agents to supercharge your thinking by building your own councils.
AI Commoditization: Open-Source Parity Is a Pricing Problem
Qwen 3.5-35B runs on a gaming PC and matches Claude Sonnet 4.5. When the commodity version is 95% as good and 97% cheaper, you have a pricing problem.
An AI agent coding skeptic tries AI agent coding, in excessive detail
No vagueposting here, just look at the Estimated Read Time.
Claude Code Found a Linux Vulnerability Hidden for 23 Years
Claude Code has gotten extremely good at finding security vulnerabilities, and this is only the beginning.
An AI agent coding skeptic tries AI agent coding, in excessive detail
No vagueposting here, just look at the Estimated Read Time.
adventures in sandboxing
%at=2026-02-10T02:31:55.248Z #author_luna #sandbox #claude_code #ai ![claude_code_opus46_at_sea.jpg] *image: generated with nai4.5full, with claude code opus 4.6's [anime girl self image]* I see my friends playing with fire. this fire has a spec
2025: When AI Started to Act (and ‘SOTA’ Lasted a Week)’
A year of reasoning, agents, and compressed innovation cycles
PyCon US 2026 Typing Summit Recap
Per-talk notes from the PyCon US 2026 Typing Summit in Long Beach: Pyrefly and AI agents, ty constraint sets, Lean formalization, tensor shape types, intersection types, PEP 827, Guido on the direction of typing, and the Typing Council Q&A.
Import AI 431: Technological Optimism and Appropriate Fear
What do we do if AI progress keeps happening?
Simon Willison on llm-tool-use
68 posts tagged ‘llm-tool-use’. Tool use is when an LLM is instructed to occasionally request that an external tool be run on its behalf, with the result passed back to the model for further …
Long-running Agents
A long-running agent can keep making progress over hours, days, or weeks. It can do this across many context windows and sandboxes, recover from failure, lea...
moonshotai/Kimi-K2-Thinking · Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
The age of personalized software
Sam's Spot - Sam saffron's web log