Field notes · AI agents, Hermes & Claude Code By Julian Goldie

Hermes Agent Prompt Caching: How v0.23.3 Warms, Compacts and Expires the Cache, and What to Set

The 20 November lifecycle controller explained: auto tiers, warming, compaction, offload and provider differences.

Get The AI Profit Stack Join AIPB →
  • 01 1,000+ done-for-you AI agent workflows
  • 02 5 live coaching calls / week with me
  • 03 7-day refund, cancel anytime
  • 04 3,400+ AI operators inside

Hermes Agent prompt caching got a single lifecycle controller in v0.23.3 on 20 November 2026, and this guide shows exactly how it warms, compacts and expires your cache, and which settings to touch.

Get the Backlink Gameplan →

Prompt caching lets the model provider reuse the start of your prompt (system prompt, skills, earlier turns) at a discounted read rate instead of reprocessing it every time.

The v0.23.3 release notes list four changes: warming through long tool runs, compaction of idle chats before the one-hour cache expires, Nous Portal on the full one-hour lifecycle, and prompt_caching.cache_ttl defaulting to auto.

This builds on v0.23.2 from 16 November, which introduced one prompt-cache clock per session.

The config block

prompt_caching:
  cache_ttl: "auto"        # default; or "5m" / "1h"; "off" disables
  route_ttl_ceilings: {}    # e.g. {nous: "5m"}

Per the configuration docs, "5m" and "1h" are sent as-is, any off value disables caching, and any other value falls back to "5m".

Caching turns on automatically when the provider supports it; there is nothing to enable.

What "auto" actually does

  • A session you type into (CLI, TUI, Desktop, Telegram, Discord, Slack and other chat platforms) starts on the 5-minute tier.
  • It moves to the 1-hour tier from the first message you send after a pause of more than five minutes, and stays there for the session.
  • /new starts on 5 minutes again.
  • Machine-paced sessions stay on 5 minutes: subagents, cron, hermes -q one-shots, webhooks, Kanban workers and the API server.
  • Delegated subagents are always clamped to 5 minutes.

The reason: on Anthropic-style caching a 1-hour write costs 2x the base input rate and a 5-minute write 1.25x, so the 1-hour tier only pays off when turns are more than five minutes apart (see Anthropic's prompt caching docs).

The Hermes docs report that on one install with parked and resumed sessions, auto cut the interactive cache-write bill by roughly 40 percent.

Step 1: keep the default

Leave cache_ttl: "auto" unless you have a reason to pin.

Pin "1h" only if every session has long gaps; pin "off" for OAuth plans that bill cache writes as extra usage or proxies that add their own cache_control.

Step 2: read the cache clock

Run /context in a session.

It prints lines like Prompt cache: warm until 14:32 (58m 10s left; 1h, proven).

Rank in AI Search or It's FREE →

"Proven" means the response named the tier; "declared" is what was requested; "assumed" or "unknown" means Hermes will not warm or compact at the cache edge.

🔥 Want Hermes tuned for long agent runs? Inside the AI Profit Boardroom you get step-by-step video trainings, the installable Agent OS, four coaching calls a week and 3,400+ members running the same agents. → Join the AI Profit Boardroom

How warming works

At five sixths of the cache lifetime (about 4m10s on 5 minutes, 25 minutes on 30, 50 minutes on 1 hour) Hermes re-sends the last request with a one-token output cap.

While the chat is idle it does this at most three times per message you send.

While a tool call is running it keeps warming until the tool returns, bounded only by cost: the keep-alives never cost more than the single full re-send they prevent.

Keep-alives need known cache prices and a cached prompt of at least 1,024 tokens, and are never sent for budget-based extended thinking.

How compaction at the cache edge works

When the idle keep-alives run out, Hermes compacts the chat if the cache window is at least 30 minutes, the context is at least half the compaction threshold and the chat grew since its last compaction.

Removed tool results are saved byte for byte under the cache/pruned folder inside your Hermes home with a pointer left in place.

Expire and offload

Otherwise the entry expires and old tool results are offloaded to disk, so the chat is already small when you come back.

Provider differences

RouteCache behaviour per the docs
Claude (Anthropic API, OpenRouter, Nous Portal)Cache markers; 5m or 1h as requested, capped by route ceiling
Nous PortalShips a 1h ceiling
Qwen CloudCapped at 5 minutes
OpenAI-compatible routesAutomatic caching, treated as 30 minutes
Responses API (openai-codex, xAI)Never warmed
Local endpoints (llama.cpp, vLLM, Ollama)Never expire on the clock

Route ceilings

Use route_ttl_ceilings to cap a provider that keeps less than it is asked for, for example {nous: "5m"}; a ceiling also sets the warming schedule.

Where the cache breakpoints go

Anthropic allows at most four cache_control breakpoints per request, and Hermes uses a strategy the developer guide calls system_and_3.

Join the AI Profit Boardroom →

  • Breakpoint 1: the system prompt, stable across all turns.
  • Breakpoints 2 to 4: the third-to-last, second-to-last and last non-system messages, as a rolling window.

The marker is {"type": "ephemeral"}, with "ttl": "1h" added for the one-hour tier.

For string content Hermes converts the message into a text block carrying the marker; for list content it adds the marker to the last element; tool messages carry it directly on native Anthropic.

What breaks the cache

Cache hits need an exact prefix match, so anything that changes earlier content invalidates everything after it.

  • Editing the system prompt mid-session: breakpoint 1 changes and the whole prefix is rewritten.
  • Inserting or removing earlier messages: the prefix no longer matches from that point on.
  • Switching model: the guide is explicit that model identity is part of the cache key, so a /model switch, a fallback to another model or a credential-pool rotation onto another account gives zero cache hits on the next request.
  • Compression: the compressed region loses its cache, but the system prompt cache survives and the rolling window rebuilds within one or two turns.

If cost spikes after a fallback fires, this is why.

Human-paced vs machine-paced sessions

Under auto, the TTL resolves once per session from the session's source.

Machine-paced sources stay on 5 minutes: subagent, cron, oneshot, webhook, kanban, api, api_server, tool, batch and curator.

Side agents started with /background, /btw or /refine also take 5 minutes, and a fork routed to a different model is clamped to 5 minutes like a delegated child.

Human-paced sessions, where you take breaks between turns, are the ones that benefit from the one-hour tier.

How compaction and caching interact

Hermes runs two compression layers, per the same guide.

The agent's own context compressor fires at 50 percent of the context window by default, inside the tool loop, using real token counts from the API.

A gateway safety net fires at 85 percent before the agent sees a message, to catch sessions that grew overnight in Telegram or Discord.

The v0.23.3 change adds cache-edge compaction on top: when idle keep-alives are spent, Hermes compacts the chat before the one-hour cache expires, so the next turn starts small instead of re-sending a cold, full history.

Get the Backlink Gameplan →

Replay size and warming cost

A keep-alive replays the last request with a tiny output cap: one token for most models, 16 for OpenAI GPT-5 and o-series models.

Warming decisions happen once at five sixths of the TTL, at least 10 seconds before expiry.

Pricing uses the catalog's one-hour write rate when the catalog lists one, otherwise a fixed 2x multiple, per the guide.

Anthropic's own prompt caching docs explain the read and write pricing behind those multiples.

Cost-saving habits

  • Pick one model per session and avoid mid-chat switches.
  • Keep skills and system prompt edits for new sessions.
  • Use /new for a new topic rather than dragging a long history along.
  • Leave cache_ttl on auto unless you have measured a reason to pin.
  • Check /context after long tool runs to confirm the cache stayed warm.

OpenAI-style routes and long retention

The developer guide splits routes into two specs.

A request carrying cache_control markers is treated as Anthropic-spec on any wire or gateway, with 5m or 1h declared as asked and clamped by the route ceiling.

Every other cached request is OpenAI-spec: declared at 30 minutes, with each read refreshing the window to the later of the existing end or read time plus 30 minutes.

A request sent with prompt_cache_retention: 24h keeps its warming deadline at 30 minutes but its estimated end at 24 hours, so it is not pruned cold on the clock; the guide notes chatgpt.com Codex strips that field, so 30 minutes stands there.

Learned routes

Hermes also learns how long each route really keeps a cache, across sessions.

It records the gap between requests and whether the cache was still there, and only adopts a learned rule after enough agreeing samples from distinct sessions.

Learned facts expire over time, so a provider change does not lock in stale behaviour.

This sits under the Hermes home directory and needs no configuration.

Verify savings in the API usage fields

Anthropic responses report cache_creation_input_tokens and cache_read_input_tokens in the usage block, per the prompt caching docs.

On a warm session most input should show up as cache reads; a turn dominated by cache creation means the prefix was rewritten.

Rank in AI Search or It's FREE →

Line those fields up against /context output to confirm the controller is doing what it reports.

Nous Portal and catalog markers

v0.23.3 puts Nous Portal on the full one-hour lifecycle, per the release notes.

The developer guide adds that a provider profile can let its /models catalog decide which models get cache markers; for Nous, a listed cache-write rate means markers are sent.

If a Nous route keeps less than it is asked for on your account, cap it with route_ttl_ceilings so warming follows the real lifetime instead of the requested one.

A working session pattern

  1. Start a session with /new and pick the model once.
  2. Load the skills you need at the start, not halfway through.
  3. Let long tool runs finish; the controller keeps the cache warm meanwhile.
  4. Step away freely; idle keep-alives and cache-edge compaction handle the gap.
  5. Start a fresh session for the next unrelated job.

That pattern lines up with how the controller is designed and gives the highest cache read share.

Troubleshooting

  • Keep-alives: off (missed twice): the cache was gone twice in a row; a /model switch restarts warming.
  • unknown: a model switch, fallback or compaction moved the cache.
  • High cache-write cost on subagents: expected, they stay on 5 minutes.

For model picks that cache well, see our best Hermes Agent model guide, and learn the full setup in the AI Profit Boardroom.

Join the AI Profit Boardroom →

FAQ: Hermes Agent Prompt Caching

Is prompt caching on by default in Hermes?

Yes. Hermes turns it on automatically when the provider supports it, and cache_ttl defaults to auto.

What does cache_ttl auto do?

Typed sessions start on the 5-minute tier and move to 1 hour after your first pause of more than five minutes; machine-paced sessions stay on 5 minutes.

How do I turn prompt caching off?

Set prompt_caching.cache_ttl to "off" in config.yaml.

Why does Hermes send tiny requests while a tool runs?

Those are keep-alives that keep the cache warm so the next turn reads the cache instead of re-sending everything.

How do I see the cache state?

Run /context; it shows when the cache stays warm and whether the tier is proven, declared or assumed.

📺 Video notes + links to the tools 👉

🎥 Learn how I make these videos 👉

🆓 Get a FREE AI Course + Community + 1,000 AI Agents 👉

About Julian

I'm Julian Goldie, an SEO entrepreneur, author and founder of the AI Profit Boardroom, which has 3,400+ members.

I help business owners scale with AI agents, automation and SEO.

  • I've built a 7-figure SEO and link building agency, Goldie Agency, with a team of 70+ SEO experts.
  • I've grown a YouTube channel to 400,000+ subscribers.
  • I wrote "Link Building Mastery".
  • My Udemy courses have taught over 29,000 students.

→ Get my best AI training inside the AI Profit Boardroom

Update to v0.23.3, leave cache_ttl on auto and check /context after your next long tool run.

For guided setups, join the AI Profit Boardroom, and keep this page as your Hermes Agent prompt caching reference.

§ Member reviews

What members say about the AI Profit Boardroom

Read the member reviews on Skool →
Ahmad Bukhari: Julian doesn't gatekeep. That's why I'd recommend this group. He's extremely active here, sharing what's working right now and jumping in to help anyone who's stuck. If he can give it away, he does. He keeps the vibe positive and pushes everyone to help each other. His challenges reward members for getting people unstuck, so the whole room shares knowledge, not just him. Want a generous, positive place to learn AI and actually use it? This is it. Thanks Julian!
Ahmad Bukhari
Julian doesn't gatekeep. That's why I'd recommend this group. He's extremely active here, sharing what's working right now and jumping in to help anyone who's stuck. If he can give it away, he does. He keeps the vibe positive and pushes everyone to help each other. His challenges reward members for getting people unstuck, so the whole room shares knowledge, not just him. Want a generous, positive place to learn AI and actually use it? This is it. Thanks Julian!

Still a paying member after 1 month

Jay Parsons: There are so many helpful members in this amazing community ready to help out with suggestions and solutions whenever someone asks a question. I am amazed at the progress I have made in learning AI. I came here to learn how to use AI better and apply what I learned to my business. I am learning so much more than I even dreamed of. I have built an Agentic OS and use it every day, and I have no background in coding. There are an incredible amount of useful tutorials on everything AI.
Jay Parsons
There are so many helpful members in this amazing community ready to help out with suggestions and solutions whenever someone asks a question. I am amazed at the progress I have made in learning AI. I came here to learn how to use AI better and apply what I learned to my business. I am learning so much more than I even dreamed of. I have built an Agentic OS and use it every day, and I have no background in coding. There are an incredible amount of useful tutorials on everything AI.

Still a paying member after 11 months

Brian O'Connor: At 47, with 30 years in sales and finance, embracing AI was exciting but overwhelming. While building a business to help others adopt AI, I found myself chasing tools and fearing I was falling behind. Julian and the AI Profit Boardroom helped me refocus through practical, personalized advice and a supportive community. I now feel more confident adapting to change while staying focused on my goals. Thank you, Julian, for bringing this community together!
Brian O'Connor
At 47, with 30 years in sales and finance, embracing AI was exciting but overwhelming. While building a business to help others adopt AI, I found myself chasing tools and fearing I was falling behind. Julian and the AI Profit Boardroom helped me refocus through practical, personalized advice and a supportive community. I now feel more confident adapting to change while staying focused on my goals. Thank you, Julian, for bringing this community together!

Still a paying member after 2 months

Neil Martin: This is a truly amazing community, thanks in no small part to Julian's prodigious output and tireless efforts to provide value to his subscribers. Nothing is too much trouble for him. If you ask a question it gets answered in short order usually with a custom video. I have been in quite a few groups and this level of service is unparalleled. If you are serious about getting an AI education, you need to be in this group!
Neil Martin
This is a truly amazing community, thanks in no small part to Julian's prodigious output and tireless efforts to provide value to his subscribers. Nothing is too much trouble for him. If you ask a question it gets answered in short order usually with a custom video. I have been in quite a few groups and this level of service is unparalleled. If you are serious about getting an AI education, you need to be in this group!

Still a paying member after 7 months

Crystal Hill: AIPB has inspired me to go beyond learning AI as individual tools, skills, and prompts, and start thinking in terms of larger systems. What I really appreciate is how responsive Julian is and how quickly the material evolves with the technology. You’re being exposed to the right tools and ideas as they emerge. AIPB has given me the confidence to learn, experiment, build, and think much bigger about what’s really possible with AI.
Crystal Hill
AIPB has inspired me to go beyond learning AI as individual tools, skills, and prompts, and start thinking in terms of larger systems. What I really appreciate is how responsive Julian is and how quickly the material evolves with the technology. You’re being exposed to the right tools and ideas as they emerge. AIPB has given me the confidence to learn, experiment, build, and think much bigger about what’s really possible with AI.

Still a paying member after 5 months

Maurice Morrison: AI Profit Board Room has been a game-changer. The insights are practical, timely, and focused on turning AI into real business results—not just theory or hype. The community is supportive, the guidance is clear, and I leave every session with ideas I can actually apply. If you’re serious about using AI to grow smarter and move faster, this is absolutely worth joining.
Maurice Morrison
AI Profit Board Room has been a game-changer. The insights are practical, timely, and focused on turning AI into real business results—not just theory or hype. The community is supportive, the guidance is clear, and I leave every session with ideas I can actually apply. If you’re serious about using AI to grow smarter and move faster, this is absolutely worth joining.

Still a paying member after 2 months

Jeremy Murdock: I have truly enjoyed my experience here. The group of folks on the calls is quite diverse and I find it comforting to know that I am not the only one going through the changes that are happening at a faster and faster pace. I have chosen to embrace the change and go full force into integrating AI and what I have learned form the group and Julian. I am also impressed and appreciative that he actually takes the time to help with challenges I have along the way and get involved personally!
Jeremy Murdock
I have truly enjoyed my experience here. The group of folks on the calls is quite diverse and I find it comforting to know that I am not the only one going through the changes that are happening at a faster and faster pace. I have chosen to embrace the change and go full force into integrating AI and what I have learned form the group and Julian. I am also impressed and appreciative that he actually takes the time to help with challenges I have along the way and get involved personally!

Still a paying member after 10 months

Michael Lovitch: I have been a part of a million ai groups in Skool for years now. They are all okay and have helped, but usually they get out of date and don't offer real service to help you through especially at the beginning. This group is different, Julian actually answers questions which is crazy! And although ai marketing moves at the speed of light, they stay current and actually help you keep up… Highly recommended!
Michael Lovitch
I have been a part of a million ai groups in Skool for years now. They are all okay and have helped, but usually they get out of date and don't offer real service to help you through especially at the beginning. This group is different, Julian actually answers questions which is crazy! And although ai marketing moves at the speed of light, they stay current and actually help you keep up… Highly recommended!

Still a paying member after 3 months

Hoebing Lo: I don't know of any other community that is as active and sharing. Julian is a star and he answers so many questions in lightning speed. If you want a great place to start with AI here it is. No question too simple. No workflow too complicated. The community is about action and results not education so if you are looking to win at AI then come for the ride of a lifetime!
Hoebing Lo
I don't know of any other community that is as active and sharing. Julian is a star and he answers so many questions in lightning speed. If you want a great place to start with AI here it is. No question too simple. No workflow too complicated. The community is about action and results not education so if you are looking to win at AI then come for the ride of a lifetime!

Still a paying member after 5 months

Darren Wagstaff: Very responsive... great personal support (not an ai chat bot). Very clear straight forward trainings in easy to understand actionable language. I'm in numerous Skool groups but this one has the most human and real community by far. No question gets ignored. You never fell like just another number. It's great to see so many members getting great results with Ai.
Darren Wagstaff
Very responsive... great personal support (not an ai chat bot). Very clear straight forward trainings in easy to understand actionable language. I'm in numerous Skool groups but this one has the most human and real community by far. No question gets ignored. You never fell like just another number. It's great to see so many members getting great results with Ai.

Still a paying member after 2 months

Rod Martin: Julian is brilliant, and more prolific than I can keep up with (which is a great thing). His AgentOS is remarkable, and he is THE go-to guy for what you can really do (and not do) with the ever-increasing deluge of new AI upgrades and tools. HIGHLY recommend.
Rod Martin
Julian is brilliant, and more prolific than I can keep up with (which is a great thing). His AgentOS is remarkable, and he is THE go-to guy for what you can really do (and not do) with the ever-increasing deluge of new AI upgrades and tools. HIGHLY recommend.

Still a paying member after 5 months

Yosuke Sato: The best Skool course you get for AI! I have attended a few Skool courses before, but AI Profit Boardroom is the most active community and has great support from the management team, including Julian! You can get the most up-to-date information and practical implementations in this community. I highly recommend this!
Yosuke Sato
The best Skool course you get for AI! I have attended a few Skool courses before, but AI Profit Boardroom is the most active community and has great support from the management team, including Julian! You can get the most up-to-date information and practical implementations in this community. I highly recommend this!

Still a paying member after 19 months

⚑ Next step

Ready to run AI agents that actually do the work?

Join 3,400+ operators inside the AI Profit Boardroom. Get 1,000+ plug-and-play AI agent workflows, 5 coaching calls a week, and a community that holds you accountable.

Join The AI Agent Community →

7-day no-questions refund · Cancel anytime

Found this useful? Get more of this site in your Google.

One tap adds us as a preferred source — Google then shows you our posts more often in Search, Top Stories, and AI Overviews.

← Back to all posts