How do I use Claude Opus 5 on Secret Chat?

Open the Secret Chat app and select the Claude model. Leave it on Smart Agentic and the router escalates hard questions to Opus 5 automatically, or pick Research in the mode selector for a deep, web-grounded answer that runs on Opus 5 directly.

Does every Claude answer on Secret Chat now use Opus 5?

No, and deliberately so. Fast Chat still runs on the quick Claude tier and Web Search on the mid tier, because sending simple questions through a frontier reasoning model would be slower and more expensive for no benefit. Opus 5 handles what the Smart Agentic router escalates plus the Research mode.

Can I turn off Claude Opus 5's reasoning?

Not in the app. Opus 5 reasons by default and decides for itself how deeply to think about each request. If you want a fast, direct answer instead, use the Fast Chat mode, which runs on the quick Claude tier.

Is Claude Opus 5 better than Claude Fable 5?

Fable 5 remains Anthropic's most capable generally available model and sits above the Opus tier, so it still wins on the very hardest reasoning tasks. Opus 5 narrows the gap considerably and is what Smart Agentic Claude escalates to automatically; Fable stays a manual choice for when you decide a task deserves it.

Does the upgrade cost me more credits?

No. Claude Opus 5 replaced Opus 4.8 at the same credit rates on Secret Chat, so an answer costs you what it did before. Because Opus 5 reasons before replying, a hard question can use more tokens than a simple one (as with any reasoning model), but the rate itself is unchanged.

Does Claude Opus 5 sit on an AI Council panel?

No, and deliberately. The AI Council seats each family's mid tier rather than its flagship, because what a council produces as evidence is models from different companies disagreeing, not each single answer being the deepest one obtainable. Claude sits on every branded panel, just not through Opus. Reach for Opus 5 or Research when you want depth from one model, and for a council when you want to see whether several companies agree on the same question.

Is Claude Opus 5 private on Secret Chat?

Secret Chat forwards your prompts through its own proxy, so you use Opus 5 without an Anthropic account and without your requests being tied to you. No profile is built, no chat is associated with your identity, and your history stays in your browser. Provider-side retention may still apply, but your query reaches the model anonymized. Note that the content of your messages is sent verbatim — only your identity is removed.

Claude Opus 5 Is Now Live on Secret Chat AI

Published July 25, 2026 · Updated August 19, 2026

Another upgrade for everyone who reaches for Claude on the hard questions: Claude Opus 5 is now live on Secret Chat AI. Anthropic's newest Opus-tier model has replaced Claude Opus 4.8 everywhere Opus was already doing the heavy lifting: the answers Smart Agentic Claude escalates to, and the Research mode you can pick yourself.

You don't have to switch anything on. If you already use Claude on Secret Chat, the deepest answers are simply coming from a stronger model as of today, at no change in what they cost you in credits. Here's what Opus 5 actually is, where it shows up in the app, and (as always) how you get it without a profile ever being built around you.

What Is Claude Opus 5?

Claude Opus 5 is the successor to Claude Opus 4.8 in Anthropic's Opus line: the tier built for complex agentic work, difficult coding, and demanding professional tasks. It's a step change over 4.8 rather than a refresh, with its biggest gains on problems that take many steps to get right. A few things worth knowing:

Reasoning is on by default. On earlier Opus models, extended thinking was something you switched on. Opus 5 thinks by default and decides for itself how deeply to think about each request — a simple question still gets a quick answer, a genuinely hard one gets real deliberation before the first word is written.
A one-million-token context window — the maximum and the default, and support for up to 128K tokens of output per response. On Secret Chat each answer is requested with a ceiling of 64,000 output tokens rather than the model's full 128K — roughly 48,000 words, far beyond any real answer. Anthropic reserves rate-limit capacity from that number the moment a request starts, whatever the reply turns out to be, so asking for the maximum on every call would throttle everyone for nothing. Long documents, long codebases and long conversations stay in view, and Anthropic reports that its instruction following and reasoning hold up across the whole window rather than fading at the far end.
Sharper on code review and bug-finding. It finds real bugs at a high rate per pass, and the extra findings tend to be genuine issues rather than noise.
Stronger vision. Charts, documents, diagrams and screenshots are a clear improvement, as is replicating a user interface from an image.
Better with documents and spreadsheets. It handles multi-sheet spreadsheets with non-trivial formulas and produces well-structured slide decks, which matters if you use Claude for work output rather than chat.

Where Opus 5 Shows Up on Secret Chat

On Secret Chat, "Claude" isn't a single model; it's Smart Claude: an automatic Smart Agentic router plus a set of modes you can choose yourself, each running the Claude tier that fits the job. Opus is the top of that ladder, and Opus is now Opus 5:

Smart Agentic — the default. A router reads each message and decides how to handle it: a fast reply for casual chat, a web search when the answer depends on current information, or a full escalation to Opus when the question is genuinely hard. That escalation now lands on Opus 5.
Research — the deep mode you pick manually. It runs on Opus with Anthropic's own web search and page-reading tools attached, so it gathers sources and then reasons over them. Also Opus 5 as of today.
Fast Chat and Web Search — unchanged. Quick questions stay on the fast Claude tier and the grounded lookup stays on the mid tier, exactly as before. Sending every throwaway message through a frontier reasoning model would make it slower and pricier for no benefit.
Claude Fable 5 — still there as the manual top-tier choice, above Opus, for the moments you decide a task deserves Anthropic's most capable model. Nothing about that mode changed; see our Fable 5 article for when it's worth reaching for.

In short: pick the Claude model, leave it on Smart Agentic or switch to Research, and the reasoning behind the answer is now Anthropic's newest Opus.

Two Honest Notes

Hard answers may take a beat longer. Because Opus 5 reasons before it replies and can't simply be told not to, a demanding question will spend a moment thinking before the answer starts arriving. That's the trade you're getting: fewer confidently wrong answers on problems that deserve care. If you want speed rather than depth, that's exactly what Fast Chat and Web Search are for.

It has extra safeguards. Like Claude Fable 5, Opus 5 ships with elevated safety measures around dual-use topics — certain sensitive cybersecurity and biology requests are declined by the model itself. That's Anthropic's behavior, not ours, and it's worth knowing before you point it at security research. Other Claude tiers and the other model families on Secret Chat are unaffected.

How to Use Claude Opus 5 on Secret Chat AI

There's nothing to install or enable; it's already live:

Open the Secret Chat app and select the Claude model in the model picker.
Leave it on Smart Agentic and let the router escalate to Opus 5 when your question needs it, or pick Research in the mode selector next to the message box when you already know the question is a hard one and want the deep, web-grounded treatment.

One practical tip that matters more with Opus 5 than with previous models: give it the whole task up front. It's built for long-horizon work and does its best when it can see the full specification, the constraints, and what "done" looks like before it starts, rather than discovering them across five follow-up messages.

Here's the full task, including the constraints and what a finished result looks like. Take the time you need, think it through before you start, and deliver the complete thing rather than an outline. If part of it's ambiguous, state the assumption you made and keep going. Here it is:

Private by Design, Whatever Claude You Pick

The model changed; the privacy model didn't. Secret Chat is a private gateway between you and the AI providers: your prompts are forwarded through our proxy, so you need no Anthropic account and your requests are never tied to you on the provider's side. We don't build a profile of you, and no chat is ever associated with your identity. Your conversation history lives locally in your own browser, not on our servers, and you can pay with crypto if you would rather not link a card to your AI use.

As we covered in Does Claude Store Your Conversations?, Anthropic's own consumer products keep your chats against your account for a long time. Reaching Opus 5 through Secret Chat lets you use the very same model as a stranger: retention may still apply at the provider, but your query arrives anonymized, not linked to your email or identity. The honest caveat, as always: Secret Chat removes you from your queries, it doesn't remove the content of your messages, so redacting identifying details before you send is still up to you. The full picture is on our Private Claude page.

And if you like having several models on tap for the same question, that's the whole idea. See why using GPT, Claude, Gemini, Grok and Perplexity together beats leaning on any single one. The AI Council is that idea built into the app. One question goes to several models from different companies at once, each answer appears the moment that model finishes writing it, and a cheap referee model then extracts the factual claims into a disagreement table — which model asserted each claim, which contradicted it, which never mentioned it — followed by a short synthesis of what at least two models agreed on and a list of things to verify before acting. You pick the size: a Duet of two models, a Trio of three, or the Quartet, which runs ChatGPT, Claude, Gemini and Grok together. Every model on a panel comes from a different company, which is the entire point, two tiers of the same family agreeing is far weaker evidence.

Worth knowing if Opus is what brought you here: panels seat each family's mid tier by design, so Opus 5 isn't one of the models on them. That isn't a compromise but the definition of the feature: a council's evidence comes from independent companies disagreeing, not from each individual answer being the deepest one obtainable. Depth from a single model is what Research and Opus 5 are for; whether several companies actually agree is what a council is for, and nothing stops you running both on the same question. Two limits we would rather state than imply: agreement is evidence, not proof — models share training data and can be wrong together, so disagreement is the reliable signal — and a council widens where your question travels, since it reaches every model on the panel plus the referee. Each request is still anonymous, but anonymization protects who you are, not what you type. The finished check is stored only on your device, and the Session Privacy Report lists every model that took part.

Conclusion

Claude Opus 5 is Anthropic's newest flagship Opus, and it now powers the hardest answers on Secret Chat — everything Smart Agentic Claude escalates, plus the Research mode. The fast tiers stay fast, Claude Fable 5 stays available as the deliberate manual upgrade, and your credits buy the same as they did yesterday.

Try it today at Secret Chat AI: the full multi-model lineup, your history in your own browser, and every query reaching the model anonymously.

Frequently Asked Questions

What is Claude Opus 5?
Claude Opus 5 is Anthropic's newest Opus-tier model and the successor to Claude Opus 4.8. It's built for complex agentic work, difficult coding and demanding professional tasks, reasons by default before answering, and has a one-million-token context window, supporting up to 128K output tokens per response — of which Secret Chat requests up to 64,000, around 48,000 words.
How do I use Claude Opus 5 on Secret Chat?
Open the Secret Chat app and select the Claude model. Leave it on Smart Agentic and the router escalates hard questions to Opus 5 automatically, or pick Research in the mode selector for a deep, web-grounded answer that runs on Opus 5 directly.
Does every Claude answer on Secret Chat now use Opus 5?
No, and deliberately so. Fast Chat still runs on the quick Claude tier and Web Search on the mid tier, because sending simple questions through a frontier reasoning model would be slower and more expensive for no benefit. Opus 5 handles what the Smart Agentic router escalates plus the Research mode.
Can I turn off Claude Opus 5's reasoning?
Not in the app. Opus 5 reasons by default and decides for itself how deeply to think about each request. If you want a fast, direct answer instead, use the Fast Chat mode, which runs on the quick Claude tier.
Is Claude Opus 5 better than Claude Fable 5?
Fable 5 remains Anthropic's most capable generally available model and sits above the Opus tier, so it still wins on the very hardest reasoning tasks. Opus 5 narrows the gap considerably and is what Smart Agentic Claude escalates to automatically; Fable stays a manual choice for when you decide a task deserves it.
Does the upgrade cost me more credits?
No. Claude Opus 5 replaced Opus 4.8 at the same credit rates on Secret Chat, so an answer costs you what it did before. Because Opus 5 reasons before replying, a hard question can use more tokens than a simple one (as with any reasoning model), but the rate itself is unchanged.
Does Claude Opus 5 sit on an AI Council panel?
No, and deliberately. The AI Council seats each family's mid tier rather than its flagship, because what a council produces as evidence is models from different companies disagreeing, not each single answer being the deepest one obtainable. Claude sits on every branded panel, just not through Opus. Reach for Opus 5 or Research when you want depth from one model, and for a council when you want to see whether several companies agree on the same question.
Is Claude Opus 5 private on Secret Chat?
Secret Chat forwards your prompts through its own proxy, so you use Opus 5 without an Anthropic account and without your requests being tied to you. No profile is built, no chat is associated with your identity, and your history stays in your browser. Provider-side retention may still apply, but your query reaches the model anonymized. Note that the content of your messages is sent verbatim — only your identity is removed.