{"playbook": "llm-cost", "title": "Controlling LLM cost", "description": "Cutting token spend and API bills: model routing, prompt and KV caching, batching, context budgets, moving work from models to code.", "url": "https://wpnews.pro/playbooks/llm-cost", "count": 6, "recommendations": [{"advice": "Cap loop context or restart sessions from a handoff file of current priorities, and prune the system prompt, since caching was 78% of the bill and pruning would save about $93 a day.", "takeaway": "In a long agent loop, caching was 78% of the Claude Code bill at API list prices because each turn re-reads the whole growing context; capping context or restarting from a handoff file and pruning the system prompt would save about $93 a day, up to about $155 with a leaner orientation.", "article_url": "https://wpnews.pro/news/agent-swarm-economics-why-caching-was-78-of-our-bill", "title": "Agent Swarm Economics: Why Caching Was 78% of Our Bill", "source_url": "https://verial.xyz/posts/agent-swarm-economics", "published_at": "2026-11-17T21:10:31.298417+00:00"}, {"advice": "Trim or summarize conversation history before each request instead of re-sending every prior turn, since a 50-message chat can send roughly 19x the input tokens of the first message.", "takeaway": "Re-sending chat history makes message 50 send about 19x the input tokens of message 1 (10,340 vs 540 in the worked example), so input cost grows with every turn.", "article_url": "https://wpnews.pro/news/why-message-50-of-your-chatbot-costs-19x-more-than-message-1", "title": "Why message 50 of your chatbot costs 19x more than message 1", "source_url": "https://dev.to/deployedlabs/why-message-50-of-your-chatbot-costs-19x-more-than-message-1-1no3", "published_at": "2026-11-17T21:01:00+00:00"}, {"advice": "Test lighter models against your own tasks before defaulting to a frontier model, and scope agent requests with check-ins; reasoning mode on Mistral Large 4 doubled tokens for no accuracy gain.", "takeaway": "Turning on reasoning mode in Mistral Large 4 more than doubled token use with no measurable accuracy gain, so teams should test whether a smaller model on a well-documented codebase suffices.", "article_url": "https://wpnews.pro/news/engineering-leaders-hire-for-judgment-as-ai-agents-take-on-more-of-the-coding", "title": "Engineering Leaders Hire For Judgment As AI Agents Take On More Of The Coding Work", "source_url": "https://www.thestateofai.com/news/ai-agents-engineering-hiring-judgment-smaller-models-david-woollard-standard-ai", "published_at": "2026-11-17T20:28:26+00:00"}, {"advice": "Weigh the 3-point index gap against per-task price when routing work: GPT-6.2 Astra scored 62 at about $4 per task versus Claude Fable 5.2's 65 at $26.54.", "takeaway": "On Artificial Analysis's index, GPT-6.2 Astra scored 62 at about $4 per task versus Claude Fable 5.2's 65 at $26.54, a 3-point gap at roughly one-sixth the cost per task.", "article_url": "https://wpnews.pro/news/openais-gpt-6-2-astra-trails-claude-fable-5-2-by-three-points-on-an-independent", "title": "OpenAI’s GPT-6.2 Astra trails Claude Fable 5.2 by three points on an independent AI index at a fraction of the cost per task", "source_url": "https://www.rdworldonline.com/openais-gpt-6-2-astra-trails-claude-fable-5-2-by-three-points-on-an-independent-ai-index-at-a-fraction-of-the-cost-per-task/", "published_at": "2026-11-17T19:36:04+00:00"}, {"advice": "Route reasoning-heavy steps to a frontier model and mechanical execution to a cheaper, faster model; Google's Orchestrator matched Carbon's success rate with 24% less quota and 19% faster completion.", "takeaway": "Routing reasoning steps to Carbon and execution steps to Hydrogen via Orchestrator matched Carbon's success rate while using 24% less quota and finishing 19% faster.", "article_url": "https://wpnews.pro/news/google-s-carbon-and-hydrogen-models-how-the-orchestrator-cuts-ai-costs", "title": "Google's Carbon and Hydrogen Models: How the Orchestrator Cuts AI Costs", "source_url": "https://www.mindstudio.ai/blog/google-gemini-carbon-hydrogen-orchestrator/", "published_at": "2026-11-17T00:00:00+00:00"}, {"advice": "Route merge-quality coding to GPT-6.2 Sol at $2/$10 and reserve Astra at $8/$40 for agent deliverables needing slides, spreadsheets or science terminal runs, since Sol trails Astra by only 0.5 points on FrontierCode 2.0.", "takeaway": "OpenAI's vendor table puts GPT-6.2 Sol within 0.5 points of Astra on FrontierCode 2.0 at roughly one-fifth the cost, and Astra 5.5 points ahead of Claude Fable 5.2 on Agents Last Exam at about 5% of the cost.", "article_url": "https://wpnews.pro/news/gpt-6-2-astra-vs-sol-vs-claude-fable-5-2-vs-gemini-4-argon-november-2026", "title": "GPT-6.2 Astra vs Sol vs Claude Fable 5.2 vs Gemini 4 Argon (November 2026)", "source_url": "https://explainx.ai/blog/gpt-6-2-astra-sol-vs-claude-fable-gemini-comparison-november-2026?utm_source=rss&utm_medium=feed&utm_campaign=blog", "published_at": "2026-11-17T00:00:00+00:00"}]}