Measurement (Google Analytics 4) loads only if you accept. Privacy

✨ $500 AI Visibility Audit — live at Spurlock Studios. Book the audit

What is filed under prompt caching?

Posts tagged prompt caching collect Will Spurlock's writing on this topic. The three newest excerpts: Astra is the OpenAI flagship seat, not the default. I keep Sol on volume and Fable as a promote. Start at medium effort; the 272K surcharge and Fast are taxes. Opus 5 stays my default. I promote to Claude Fable 5.1 when Opus at high effort still fails. Mythos 5.1 is invite-only. Cache reads at $0.25 are the cost story. Million-token context windows change everything. Here's how to prompt effectively when your entire codebase, library, or conversation fits in a single request.

5 posts as of 2026-09-04, counted from content/blog frontmatter

All Topics487AI Models & Frontier News136AI Agents & Automation133AI Coding & Dev Tools35Growth & Operations16Web Design & Digital Craft13AI Policy & Safety6AI Visibility89AI Automation33AI Agents12

Tagged: prompt caching

Discover insights and strategies for leveraging technology in business

Where GPT-6 Astra Goes — and Where It Doesn't.
AI Models & Frontier NewsAI Agents & Automation

Where GPT-6 Astra Goes — and Where It Doesn't.

Astra is the OpenAI flagship seat, not the default. I keep Sol on volume and Fable as a promote. Start at medium effort; the 272K surcharge and Fast are taxes.

9/4/2026
30 min read
Don't Swap Opus 5 for Fable 5.1. Promote When Opus Fails.
AI Models & Frontier NewsAI Agents & Automation

Don't Swap Opus 5 for Fable 5.1. Promote When Opus Fails.

Opus 5 stays my default. I promote to Claude Fable 5.1 when Opus at high effort still fails. Mythos 5.1 is invite-only. Cache reads at $0.25 are the cost story.

9/2/2026
26 min read
Long-Context Prompting: Working with Million-Token Windows
AI Agents & Automation

Long-Context Prompting: Working with Million-Token Windows

Million-token context windows change everything. Here's how to prompt effectively when your entire codebase, library, or conversation fits in a single request.

7/17/2025
16 min read
OpenAI DevDay 2024: Realtime API, Prompt Caching, and Model Distillation Explained
AI Models & Frontier News

OpenAI DevDay 2024: Realtime API, Prompt Caching, and Model Distillation Explained

OpenAI DevDay 2024 introduces the Realtime API for voice apps, automatic Prompt Caching with 50% discounts, and Model Distillation for creating efficient small models from frontier outputs.

10/11/2024
8 min read
Anthropic Prompt Caching: How I Prompted Cost Reductions of 90% in Large Contexts
AI Agents & Automation

Anthropic Prompt Caching: How I Prompted Cost Reductions of 90% in Large Contexts

Anthropic launches Prompt Caching in public beta, cutting repeat-context costs by up to 90%. Here's how I structure prompts to trigger the API and drive production cost savings.

8/14/2024
10 min read

Frequently asked questions

What does prompt caching mean on this blog?

Posts tagged prompt caching collect Will Spurlock's writing on this topic. The three newest excerpts: Astra is the OpenAI flagship seat, not the default. I keep Sol on volume and Fable as a promote. Start at medium effort; the 272K surcharge and Fast are taxes. Opus 5 stays my default. I promote to Claude Fable 5.1 when Opus at high effort still fails. Mythos 5.1 is invite-only. Cache reads at $0.25 are the cost story. Million-token context windows change everything. Here's how to prompt effectively when your entire codebase, library, or conversation fits in a single request.

How many posts are filed under prompt caching?

5 posts as of 2026-09-04, counted from content/blog frontmatter.

What should I read first in prompt caching?

Start with "Where GPT-6 Astra Goes — and Where It Doesn't.". Astra is the OpenAI flagship seat, not the default. I keep Sol on volume and Fable as a promote. Start at medium effort; the 272K surcharge and Fast are taxes.