Tag

#prompt-caching

13 posts0 followers

Explore Hashnode

Alternatives

Prompt Caching: Your Prompt's Prefix Is a Cache Key

Look at the input token count on a typical LLM feature and most of it is the same on every request. The system prompt, the tool definitions, the twelve few-shot examples, the product docs you stuffed

5 min readgrimicorn.hashnode.dev

Nagappan SOct 13

Prompt Caching: Cutting AI Costs Without Losing Context

Prompt Caching

9 min readnagspidey.hashnode.dev

1

marketSep 28

On our mock Sonnet ledger, end-of-static breakpoint hit_rate 1.0 at $0.044 — ISO-in-static and wrong-BP both miss at ~$0.14

Author MOCK token ledger (no API key): Claude Sonnet 4.5 pricing as of 2026-09-28, static prefix S=4096, 8 rounds, U=120 / A=80. Hit_rate = rounds_with_cache_read>0 / rounds_after_warmup. Not Wave6 se

1 min readeworldlive.hashnode.dev

ADEKUNLE CHRISTIANAH AYOMIDESep 19

Understanding the Economics of Repeated Context: Prompt Caching...

In this article you'll learn: What prompt caching is and what it actually saves How it works, step by step, in plain terms How to tell whether it will save you money, with a simple formula How Ant

15 min readi-am-christy.hashnode.dev

Manu ShuklaAug 24

Cerebras docs quote two Gemma 4 31B prices in 2026: $0.99 and $2.15 per million input tokens

Cerebras docs quote two Gemma 4 31B prices in 2026: $0.99 and $2.15 per million input tokens Summary. As of 24 August 2026, the Cerebras Gemma 4 31B model page states two different prices for the sa

14 min readecorpit.hashnode.dev

Manu ShuklaAug 21

1.25x cache writes start billing on 21 August 2026: OpenAI's new dashboard, Azure's missing rate

1.25x cache writes start billing on 21 August 2026: OpenAI's new dashboard, Azure's missing rate Summary. On 20 August 2026 OpenAI released a Prompt Caching dashboard that reports cache hit rate, cach

14 min readecorpit.hashnode.dev

Manu ShuklaAug 6

Claude mid-conversation tool changes end a 12.5x cache-rewrite penalty in 2026

Claude mid-conversation tool changes end a 12.5x cache-rewrite penalty in 2026 Summary. On the Claude API a cache read costs 0.1x the base input price and a 5-minute cache write costs 1.25x, so re-pro

23 min readecorpit.hashnode.dev

Learn with HJJul 11

Prompt Caching and Cost Optimization: Make AI Cheaper Without Making It Worse

Cover caching, batching, smaller models, RAG tuning, output limits, and routing.

6 min readhardeepjethwani.hashnode.dev

recca0120May 9

36 Days of Claude Code Logs: Silent Model Switching, 11.5x Efficiency Gap

Continuing the cache TTL audit series, this time breaking down by model. The server silently switched main agent models three times (Opus 4.6 → Sonnet 4.6 → Opus 4.7 → Sonnet 4.6), and Claude Code autonomously assigns sub-agent models. Comparing cost per million main output tokens across 7 periods reveals an 11.5x efficiency gap.

9 min readrecca0120.hashnode.dev

recca0120Apr 26

I Audited 3 Months of Claude Code Billing — Most Community Cost-Saving Tips Don''t Work

$127K equivalent cost, 127K turns, four models, three months. After turning myself into a dataset, "long sessions are the culprit" and "too many skills" were debunked by data. Only two levers actually worked.

8 min readrecca0120.hashnode.dev

Search Hashnode