optimizing-for-cost-and-intelligence changedabout-claude/models/optimizing-for-cost-and-intelligence
Nearest release: v2.1.318, published 4 hours before this site recorded the change. Shown because the two are within 24 hours of each other. Nothing here says the release caused the edit.
Recorded here
Lines+5added
Lines−3removed
From line
49
where the diff opens
First seen
14 Aug 2026
this site's first read of the page
Recorded edits33to this page, all time
The whole hunk
from line 49, old and new numbered
/
from line 49
4949
5050#### Why caching comes first
5151
52Turn on [prompt caching](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) before any other lever, because every turn of an agentic task resends the entire growing conversation: system prompt, tool definitions, and every prior turn.
52Turn on [prompt caching](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) before any other lever, because by default every turn of an agentic task resends the entire growing conversation: system prompt, tool definitions, and every prior turn.
5353
5454* A 40-turn task sends its first turn 40 times, so task cost grows with roughly the square of turn count.
5555* Caching does not stop the resending, but each resend costs about a tenth as much (a twentieth on Claude Opus 5.5 and Claude Sonnet 5.5, a fortieth on Claude Fable 5.1 and Claude Mythos 5.1) and processes faster.
5656* The prefix is billed at the [cache-read rate](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#pricing) instead of the input price, and each turn pays the 1.25x cache-write rate only for what is new.
5757
58[Message threads](https://platform.claude.com/docs/en/build-with-claude/message-threads), in beta, stop the resending while a thread lasts, until 5 minutes after its last request. They lower time to first token, not cost: the model still reads the whole conversation, so keep `cache_control` on every request. Threads don’t support compaction or most context editing.
59
5860**What good looks like.**
5961
6062* Over a full day of real traffic, agent loops read a median 84% of their input from the cache, and the top 10% of them, coding or not, read 94% or more<sup>[17](https://platform.claude.com/docs/en/about-claude/models/optimizing-for-cost-and-intelligence#refs)</sup>.
from line 170
168170 jq '.max_tokens = 0 | del(.stream)' last_request.json | \
169171 curl https://api.anthropic.com/v1/messages \
170172 -H "x-api-key: $ANTHROPIC_API_KEY" \
171 -H "anthropic-version: 2023-06-01" \
173 -H "anthropic-version: 2026-10-21" \
172174 -H "content-type: application/json" \
173175 --data-binary @-
174176 ```
from line 314
312314Route every request no one is waiting on through the [Batch API](https://platform.claude.com/docs/en/build-with-claude/batch-processing), and keep the interactive path for the rest. The Batch API takes 50% off every token of a request, including cached ones, in exchange for results arriving any time within 24 hours.
313315
314316* Batching is the second-largest free lever after caching for unattended agent work: evaluation runs, backfills, and scheduled jobs such as a recurring run of the issue-triage agent from the [token-trimming measurement](https://platform.claude.com/docs/en/about-claude/models/optimizing-for-cost-and-intelligence#trim-input-and-context-tokens).
315* Batching combines with most of this page, but not with [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode) or Claude Managed Agents sessions (see [Claude Managed Agents pricing](https://platform.claude.com/docs/en/about-claude/pricing#claude-managed-agents-pricing)).
317* Batching combines with most of this page, but not with [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode), message threads, or Claude Managed Agents sessions (see [Claude Managed Agents pricing](https://platform.claude.com/docs/en/about-claude/pricing#claude-managed-agents-pricing)).
316318
317319### Audit prompts against the current model
318320
No line in this hunk matches that.