Ask Otto, Claude Opus 5 and seven new models

[Chat Thing](https://chatthing.ai/)

Open main menu

- Features
- Use Cases
- [Pricing](https://chatthing.ai/pricing)
- Resources

[Log in](https://app.chatthing.ai/app/auth/login) [ Book a demo](https://chatthing.ai/demo) [Sign up](https://app.chatthing.ai/app/auth/register)

[← Back to blog](https://chatthing.ai/blog)Changelog # Otto now lives inside every agent, plus seven new models

![Zef](https://res.cloudinary.com/djyjvrw5u/image/upload/v1710941716/IMG_2278_d3b5e4fa69.jpg)

Zef

30 Jul 2026

~ 6 min read

---

![Otto now lives inside every agent, plus seven new models](https://res.cloudinary.com/djyjvrw5u/image/upload/v1785426041/hero_2195c49528.png)

---

### On this page

1. [Ask Otto: change your agent by describing the change](#ask-otto-change-your-agent-by-describing-the-change)
   1. [The limits](#the-limits)
   2. [Otto building agents from scratch, still in testing](#otto-building-agents-from-scratch-still-in-testing)
2. [Seven new models](#seven-new-models)
3. [Twelve retired models](#twelve-retired-models)
4. [Embed and SDK](#embed-and-sdk)
5. [Try it](#try-it)

Two weeks since the last update, and the main change is Otto. Our assistant now sits inside every agent, so you can change things by describing what you want rather than hunting through settings forms.

There's also seven new models, four new embed capabilities, and twelve models we've retired.

## Ask Otto: change your agent by describing the change

We've worked hard to make Chat Thing simple, and mostly it is. But to get the _most_ out of it you still need a certain amount of technical know-how: how context works, what the [retrieval settings](https://chatthing.ai/docs/bot-advanced-settings) actually do, [which model suits which job](https://chatthing.ai/docs/available-models). People would get that wrong. Not catastrophically, just enough that their agent was noticeably worse than it should have been. Whenever we set an agent up for someone ourselves, it performed a lot better.

Which makes us a consultancy, and we'd rather ship a product that gets you there without us.

So we taught Otto how we do it. It knows the [best practices](https://chatthing.ai/docs/prompt-engineering-tips), understands what every setting does and when to reach for it, and can talk to your agent on your behalf to check whether a change worked. The same judgement we'd apply, without us being in the room.

You can tell it what's wrong in plain language. _"It's doing this weird thing, how do we fix it?"_ is a valid prompt.

Every page inside an agent now has an **Ask Otto** button. It opens a panel beside the dashboard rather than over it, so the page stays visible and usable while you talk.

![The Ask Otto button in an agent's page header, alongside Embed, Test and Open bot](https://res.cloudinary.com/djyjvrw5u/image/upload/v1785177627/otto_trigger_e83049d211.png)

Things people ask it:

- _"Add our changelog page as a data source"_
- _"Make the tone more formal"_
- _"Set up escalation to a human when someone asks for refunds"_
- _"Review this agent's setup and suggest improvements"_

![Otto's panel open beside the agent dashboard, showing a greeting and three suggested questions](https://res.cloudinary.com/djyjvrw5u/image/upload/v1785177630/otto_panel_baaf1bfbe1.png)

When Otto changes something, the page behind it updates straight away. Rename your agent and you'll watch the new name land in the header, the breadcrumb and the open settings form while you're still typing.

![After asking Otto to rename the agent, the page header, breadcrumb and the name field all show the new name](https://res.cloudinary.com/djyjvrw5u/image/upload/v1785177631/otto_live_refresh_b57d0dd43c.png)

The one we use most is diagnosis. When a customer gets a bad answer, open the conversation in [**Conversations**](https://chatthing.ai/docs/my-chats) and hit **Ask Otto about this chat**. Otto reads the real exchange, including which sources the answer drew on, works out whether the cause was missing knowledge, retrieval settings that were too strict, the system message or the model, then proposes a fix and applies it once you agree.

It can also [talk to your agent](https://chatthing.ai/docs/bot-testing) to check the fix landed. Those test conversations are quarantined, so they don't appear in your **Conversations** list and they don't fire webhooks.

### The limits

Otto's test chats don't execute [power-ups](https://chatthing.ai/docs/power-up-basics). Otto should never fire a real [escalation email](https://chatthing.ai/docs/talk-to-a-human) or hit your API while it's poking around, so a power-up fix can't be confirmed by Otto's testing. It will configure the power-up and tell you what it changed, and you should verify the side effect yourself in a real conversation.

That's a limitation for now rather than a permanent one. We want Otto to be able to test power-ups properly, and we'll get there once we can do it without risking a real send.

A few other things stay in the dashboard on purpose. Otto won't connect Slack, Discord, WhatsApp, Telegram or email channels. It will never ask you to paste an API key, token or password into the chat. File uploads and the Notion and Google source pickers stay on the **Data sources** tab, and billing and team settings are out of scope.

It's scoped, too. Otto working inside an agent can only change that agent, not a sibling in your team, and not even if a customer message in a conversation it's reading appears to ask it to. It's available to team owners and admins, since its changes need write access.

Full walkthrough is in [the Otto docs](https://chatthing.ai/docs/ask-otto).

### Otto building agents from scratch, still in testing

The other half of this is Otto setting up a brand new agent with you. You describe what you want, or just hand it your website, and it researches your business, writes the system message, finds your pages and adds them as knowledge, matches the widget to your branding and suggests power-ups that fit.

We've been running that as a test with a portion of new signups and it's going well. Well enough that we're confident in the approach, which is why the management side has now gone out to everyone.

It is still a test though, so not every new account gets it yet, and we're still tuning. What shipped this fortnight is the management side, live for everyone. Otto creating agents from scratch is still rolling out.

## Seven new models

We diffed the entire OpenRouter catalogue against [everything we carry](https://chatthing.ai/models). Most recent releases were already covered, including [Kimi K3](https://chatthing.ai/blog/kimi-k3-lands-in-chat-thing) and [the GPT-5.6 family](https://chatthing.ai/blog/gpt-5-6-and-thirteen-new-models). These were the real gaps.

| Model | Context | Input multiplier | Output multiplier |
| --- | --- | --- | --- |
| Claude Opus 5 | 1M | 10 | 50 |
| Claude Opus 5 (Fast) | 1M | 20 | 100 |
| Gemini 3.6 Flash | 1M | 3 | 15 |
| Gemini 3.5 Flash Lite | 1M | 0.6 | 5 |
| Mistral Large 3 | 256k | 1 | 3 |
| Command A | 256k | 5 | 20 |
| Qwen3.7 Flash | 1M | 0.06 | 0.26 |

Claude Opus 5 is the one most people were waiting for, since the newest Opus you could pick was [4.8](https://chatthing.ai/blog/claude-fable-5-opus-4-8-and-a-new-model-picker). Both the standard and Fast endpoints are here with a 1M context window each, matching the pairing we already ship for 4.7 and 4.8.

Qwen3.7 Flash is the other one worth a look. At 0.06 in and 0.26 out it's an order of magnitude cheaper per token than anything else we carry, with a 1M context window and vision. Qwen joins as a new provider on the back of it.

Two caveats. Command A doesn't support power-ups, because OpenRouter advertises no tool support on that endpoint, so it won't appear as an option for agents that use them. And we haven't moved the recommended badges: Opus 4.8 and Sonnet 4.6 are still the nudged picks, and Gemini 3.5 Flash keeps its badge while 3.6 arrives unbadged. Promoting a new generation is a separate call from making it available.

## Twelve retired models

We've retired twelve models that OpenRouter no longer serves. If one of your agents was on one, it has already been moved and there's nothing for you to do.

Sixty-seven agents across 26 teams were affected. None had been used in the past 30 days, so nobody hit the failure before we caught it. Each has been moved to the newest live model in the same family, so a Gemini agent stays on Gemini.

Two constraints on that migration. Same-generation variants like (Fast), (Thinking) and (Preview) are excluded as targets, because "Claude Opus 5 (Fast)" costs double "Claude Opus 5" and picking it would have quietly doubled what every migrated agent burns. And the migration respects your plan, so a team without model selection lands on the default rather than a model it can't pick.

Finding retirements used to mean running a live completion against every model in our table, which was slow and expensive enough that nobody ran it on a schedule. There's now a read-only check that compares our list against the OpenRouter catalogue in seconds, and it can gate a release.

## Embed and SDK

Four additive capabilities for anyone [embedding an agent](https://chatthing.ai/docs/sdk) on their own site.

If you run a [multilingual site](https://chatthing.ai/blog/multilingual-your-agent-speaks-50-languages), you can now override the greeting and the example questions per locale at init, without touching your saved config:

```
window.chatThingConfig = {
  locale: "es",
  firstMessageOverride: "¡Hola! ¿En qué puedo ayudarte?",
  exampleQuestionsOverride: ["¿Horarios?", "¿Envíos internacionales?"],
};

```

The host page can also read the conversation and the loading state now, where previously the iframe only accepted instructions and never published anything back:

```
window.chatThing.onMessage((m) => console.log(m.role, m.message));
window.chatThing.onLoading((loading) => toggleSpinner(loading));
const transcript = window.chatThing.getMessages();

```

Useful for analytics, your own loading UI, or reacting to what a visitor just asked. Assistant messages arrive complete rather than token by token, so `onMessage` gives you finished text.

## Try it

Open an agent, hit **Ask Otto**, and ask it why your worst recent answer was bad. It reads the actual conversation, so it tends to be more specific than you'd expect.

[Log in and have a go](https://app.chatthing.ai/app), or [start free](https://app.chatthing.ai/app/auth/register) if you haven't yet.

### Related articles

[![Kimi K3 is live in Chat Thing: what it is, and is it actually good?](https://res.cloudinary.com/djyjvrw5u/image/upload/v1784238623/hero_a9dfc94c16.png)](https://chatthing.ai/blog/kimi-k3-lands-in-chat-thing) [<h4>Kimi K3 is live in Chat Thing: what it is, and is it actually good?</h4>](https://chatthing.ai/blog/kimi-k3-lands-in-chat-thing)

Moonshot AI's Kimi K3 launched today and it's already live in Chat Thing. Here's what Kimi K3 is, the benchmarks behind the hype, and how it performed when we ran it head-to-head against Claude Sonnet 5, GPT-5.6 and Gemini on real support-bot prompts.

[![GPT-5.6 lands in Chat Thing, plus twelve more new models](https://res.cloudinary.com/djyjvrw5u/image/upload/v1783942921/hero_c5c962c5c0.png)](https://chatthing.ai/blog/gpt-5-6-and-thirteen-new-models) [<h4>GPT-5.6 lands in Chat Thing, plus twelve more new models</h4>](https://chatthing.ai/blog/gpt-5-6-and-thirteen-new-models)

GPT-5.6 (Sol, Terra and Luna) is now live in Chat Thing, alongside twelve more new models. A 1.05M-token context window, vision and power-up support across the board, and it is unusually light on message tokens for how capable it is.

[![Claude Fable 5, Opus 4.8, and a brand-new model picker](https://res.cloudinary.com/djyjvrw5u/image/upload/v1781277929/new_models_blog_post_og_image_60c61ed10a.jpg)](https://chatthing.ai/blog/claude-fable-5-opus-4-8-and-a-new-model-picker) [<h4>Claude Fable 5, Opus 4.8, and a brand-new model picker</h4>](https://chatthing.ai/blog/claude-fable-5-opus-4-8-and-a-new-model-picker)

Claude Fable 5, the most capable model Anthropic have made, is live on Chat Thing, along with a stack of other new models and a rebuilt picker. A heads-up on what Fable costs, and which plans get to choose.

<dl>

<dt>Previous</dt>
<dd>[← Kimi K3 is live in Chat Thing: what it is, and is it actually good?](https://chatthing.ai/blog/kimi-k3-lands-in-chat-thing)</dd>

</dl>

## Footer

### Ready to answer every question, day or night?

Add an AI support agent to your site in minutes.

[Get started free](https://app.chatthing.ai/app/auth/register) [Book a demo](https://chatthing.ai/demo)

AI customer support that learns from your content and answers 24/7.

### Features

- [Channels: Deploy your bot](https://chatthing.ai/pages/features/channels)
- [Data sources: Add knowledge](https://chatthing.ai/pages/features/data-sources)
- [Power-ups: Tool calling](https://chatthing.ai/pages/features/power-ups)
- [AI Models: GPT-4, Claude, & more](https://chatthing.ai/pages/features/models)
- [SDK: Create custom assistants](https://chatthing.ai/pages/features/chatbot-sdk)
- [Bot evaluations: Test your bot](https://chatthing.ai/pages/features/testing)
- [Analytics: Topics, sentiment & insights](https://chatthing.ai/pages/features/analytics)
- [Bot Gallery: Example bots](https://chatthing.ai/gallery)
- [All features](https://chatthing.ai/features)

### Use cases

- [AI Customer Support Bots](https://chatthing.ai/pages/use-cases/customer-support)
- [AI Internal Knowledge Bots](https://chatthing.ai/pages/use-cases/internal-knowledge)
- [AI Bots for Marketing](https://chatthing.ai/pages/use-cases/marketing)
- [SDK: Create custom assistants](https://chatthing.ai/pages/features/chatbot-sdk)
- [All features](https://chatthing.ai/features)

### Channels

- [AI support for your website](https://chatthing.ai/pages/features/channels/website)
- [AI support for Slack](https://chatthing.ai/pages/features/channels/slack)
- [AI support for Discord](https://chatthing.ai/pages/features/channels/discord)
- [AI support for WhatsApp](https://chatthing.ai/pages/features/channels/whatsapp)
- [AI support for Telegram](https://chatthing.ai/pages/features/channels/telegram)
- [AI support API](https://chatthing.ai/pages/features/channels/api)

### Data sources

- [Chat with your website](https://chatthing.ai/pages/data-source/web)
- [Chat with your files](https://chatthing.ai/pages/data-source/files)
- [Notion chatbots](https://chatthing.ai/pages/data-source/notion)
- [Chat with your youtube](https://chatthing.ai/pages/data-source/youtube)
- [Chat with your RSS feed](https://chatthing.ai/pages/data-source/rss)

### Company

- [Discord](https://discord.com/invite/DGCCR34Fsj)
- [Docs](https://chatthing.ai/docs/getting-started)
- [Blog](https://chatthing.ai/blog)
- [FAQ](https://chatthing.ai/faq)
- [Status](https://status.chatthing.ai/)
- [Testimonials](https://senja.io/p/chat-thing/iq1tWz)
- [Custom AI development](https://chatthing.ai/pages/ai-development-services)

### Free tools

- [Agent Readiness Checker](https://chatthing.ai/tools/agent-readiness-checker)
- [JSON-LD Generator + Checker](https://chatthing.ai/tools/json-ld-generator)
- [llms.txt Generator & Validator](https://chatthing.ai/tools/llms-txt-generator)
- [Support Ticket Deflection Calculator](https://chatthing.ai/tools/ticket-deflection-calculator)
- [Knowledge Base Grader](https://chatthing.ai/tools/knowledge-base-grader)
- [AI FAQ Generator](https://chatthing.ai/tools/faq-generator)
- [AI Canned Response Generator](https://chatthing.ai/tools/canned-response-generator)
- [Support Bot Prompt Generator](https://chatthing.ai/tools/support-bot-prompt-generator)
- [All free tools](https://chatthing.ai/tools)

### Compare

- [Intercom alternative](https://chatthing.ai/pages/alternatives/intercom)
- [Zendesk alternative](https://chatthing.ai/pages/alternatives/zendesk)
- [Tidio alternative](https://chatthing.ai/pages/alternatives/tidio)
- [Crisp alternative](https://chatthing.ai/pages/alternatives/crisp)
- [Chatbase alternative](https://chatthing.ai/pages/alternatives/chatbase)
- [DocsBot alternative](https://chatthing.ai/pages/alternatives/docsbot)
- [ResolveAI alternative](https://chatthing.ai/pages/alternatives/resolveai)
- [ChatBotKit alternative](https://chatthing.ai/pages/alternatives/chatbotkit)
- [CustomGPT alternative](https://chatthing.ai/pages/alternatives/customgpt)
- [Botsonic alternative](https://chatthing.ai/pages/alternatives/botsonic)
- [Ada alternative](https://chatthing.ai/pages/alternatives/ada)

### Join our newsletter

Be the first to know about new features and updates.

Join

[Privacy](https://chatthing.ai/legal/privacy) [Terms](https://chatthing.ai/legal/terms) [DPA](https://chatthing.ai/legal/dpa) [GDPR](https://chatthing.ai/legal/gdpr) [Sub-Processors](https://chatthing.ai/legal/sub-processors)

[Twitter](https://twitter.com/pixelhopio) [Discord](https://discord.com/invite/DGCCR34Fsj) [YouTube](https://www.youtube.com/@ChatThingAI)

© 2026 Chat Thing. All rights reserved. Company number: 14789358. Built with  by [Pixelhop](https://pixelhop.io?utm_source=chat+thing&utm_medium=website)

## Cookies

We use cookies to run the site and understand how it's used. You can change your choice anytime.

Accept

Decline

Customise