# Agent API: prompt caching for openai/gpt-6.1-sol (cache writes billed, breakpoints dropped)

**URL:** <https://community.perplexity.ai/t/agent-api-prompt-caching-for-openai-gpt-6-1-sol-cache-writes-billed-breakpoints-dropped/6390>\
**Category:** General\
**Tags:** agent-api, api\
**Created:** [October 20, 2026, 1:06pm UTC](https://community.perplexity.ai/t/agent-api-prompt-caching-for-openai-gpt-6-1-sol-cache-writes-billed-breakpoints-dropped/6390 "2026-10-20T13:06:13Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Marius](https://sea1.discourse-cdn.com/flex001/user_avatar/community.perplexity.ai/marius/32/3724_2.png) [@Marius](https://community.perplexity.ai/u/Marius)\
**Post date:** [October 20, 2026, 1:06pm UTC](https://community.perplexity.ai/t/agent-api-prompt-caching-for-openai-gpt-6-1-sol-cache-writes-billed-breakpoints-dropped/6390/1 "2026-10-20T13:06:13Z")

</div>

Hello Perplexity API team,  
we use the Agent API (POST /v1/agent) with openai/gpt-6.1-sol for long  
structured-output calls (60–90k input tokens). While checking our costs we  
found three things about prompt caching that the docs do not cover. Test  
requests from 2026-10-20, about 10:55–10:59 UTC, are listed below.

1. Cache writes are billed but not documented  
Every gpt-6.1-sol request bills the full input as  
cache\_creation\_input\_tokens (cost.cache\_creation\_cost), which works out to  
1.25x the input price ($2.50 instead of $2.00 per 1M tokens). The Agent API  
Models page lists input, output and cache-read prices only, with no cache-write  
price. Across 215 production-like requests we were billed for 5.29M written  
tokens and 1,039 read tokens.

- Is the cache-write surcharge intended?
- Can we opt out of cache writes for requests that will not be reused?  
Example: resp\_2ea3584b-e620-4b9c-83b5-252315d591ec

1. Explicit cache breakpoints are dropped  
OpenAI documents prompt\_cache\_options (mode “explicit”) and  
prompt\_cache\_breakpoint on input\_text parts for GPT-5.6 and later. The Agent  
API accepts both without an error, but the response echoes  
prompt\_cache\_options as null and still writes the full input. Mode “explicit”  
with no breakpoint should write nothing, but it still writes everything.  
Because of this, a follow-up request that resends the same ~60–90k-token  
prefix plus new turns (our validation retries) never reads from the cache.

- Do you plan to pass OpenAI’s explicit cache breakpoints through?
- If not, is there another way to get a cache read on a shared prefix when  
new turns are appended?  
Examples:  
resp\_d9f8439a-7fce-4d0b-94f7-a406935fbb0b (explicit breakpoint, 0 read)  
resp\_1c3168c9-f86a-48fa-a735-3e41f9097a38 (same prefix + 2 turns, 0 read)

1. prompt\_cache\_key for OpenAI models  
Identical requests without prompt\_cache\_key never read from the cache (0 of  
3). With prompt\_cache\_key they did (3 of 3, about 95% cheaper). The field is  
not in the API reference schema; it appears only in the changelog under  
presets.

- Is prompt\_cache\_key the supported way to get cache reads for OpenAI models?
- Does it affect routing?
- Are there limits we should know about, such as the number of keys or  
requests per key?  
Examples:  
resp\_27f8e738-58e4-4c44-a7d1-1de857c37331 (no key, 0 read)  
resp\_d1a9796f-76f6-4f1c-bddd-792645d6704a (with key, 11,926 tokens read)  
Thank you,  
Marius