Published January 14, 2026

Anthropic Labs: Try Claude's New Experimental AI Features

Anthropic Launches Labs – A Sandbox for Experimenting with Claude's New Capabilities

Anthropic has opened access to experimental Claude features via its new Labs platform, where capabilities can be tested before their official release.

Event Source: Anthropic Reading Time: 3 – 5 minutes

Anthropic has launched a new platform called Labs. In short, it is a space where you can try out Claude's experimental capabilities before they become part of the main product.

What is Anthropic Labs and Its Purpose

What Labs Is and Why It Matters 🧪

The idea is simple: developers are constantly testing new features for Claude, but not all of them make it into the final version immediately. Previously, these developments stayed within the company. Now, Anthropic has decided to show them to users in advance – to get feedback and understand what is actually useful, and what needs more work or should be shelved.

Labs functions as a separate section on the Claude website. Features that aren't yet ready for a mass launch, but are stable enough for testing, are posted there. You can try them out, rate them, and send feedback – and this input directly influences which functions eventually become part of the product.

Extended Thinking: The First Experimental Feature

The First Feature: Extended Thinking

Along with the launch of Labs, Anthropic immediately added its first experimental capability: Extended Thinking. The gist is that before answering, the model can explicitly “think” – formulating internal reasoning that is usually hidden from the user.

Simply put, Claude reveals its thought process. This helps in tasks where accuracy and step-by-step logic are key: programming, math, and complex data analysis. You see not only the final answer, but also how the model reached it – what intermediate conclusions it drew, what it took into account, and where it might have made a mistake.

Such transparency is useful for debugging: if the model made an error, you can understand at what stage and why. Or verify that the logic is sound, even if the result seems unexpected at first glance.

How Extended Thinking Works on Anthropic Labs

How It Works in Practice

Extended Thinking is available in the free version of Claude with some restrictions, and in paid plans (Pro and Team) with broader access. To use the feature, you need to go to the Labs section on the website and activate it in the chat settings.

After that, when responding to a complex query, Claude can unfold its reasoning. This takes more time and tokens than a standard response, but gives you more control and understanding in return.

Why Anthropic Shows Unfinished Features in Labs

Why Show Unfinished Features at All?

Previously, companies more often operated according to the “make → release → observe reaction” scheme. Now, more and more projects are moving to a more open model, especially in the field of AI. The reason is simple: technologies are developing rapidly, and guessing what exactly will prove useful is becoming harder.

Labs allows Anthropic to test hypotheses on real users without risking the stability of the main product. If a feature turns out to be in demand – it will be refined and added to Claude. If not – they can calmly abandon the idea or rethink the approach.

For users, this is an opportunity to influence the development of the tool they use. And at the same time – to gain access to new capabilities before others.

What's Next for Anthropic Labs Features

What's Next

Anthropic isn't specifying exactly which features will appear in Labs in the future. But judging by the platform's logic, all sorts of experiments will end up there: from improvements in text processing to new modes of working with context or integrations.

An important point: Labs is not a beta version of Claude. The main product remains stable, and experimental features reside separately. If something breaks or doesn't work as expected, it won't affect the normal use of the model.

Overall, the approach looks logical: let people try something new, gather feedback, and refine based on it. And not pretend that everything is already perfect when it isn't.

#event #applied analysis #ai development #products #interfaces #open technologies #development_tools
Original Title: Introducing Labs
Publication Date: Jan 13, 2026
Anthropic www.anthropic.com A U.S.-based company developing large language models with a focus on AI safety and alignment.
Previous Article Clinical AI in 2026: Quieter Demos, More Real-World Practice Next Article How to Simplify Running ONNX Models on Windows with WinML

From Source to Analysis

How This Text Was Created

This material is not a direct retelling of the original publication. First, the news item itself was selected as an event important for understanding AI development. Then a processing framework was set: what needs clarification, what context to add, and where to place emphasis. This allowed us to turn a single announcement or update into a coherent and meaningful analysis.

Neural Networks Involved in the Process

We openly show which models were used at different stages of processing. Each performed its own role — analyzing the source, rewriting, fact-checking, and visual interpretation. This approach maintains transparency and clearly demonstrates how technologies participated in creating the material.

1.
Claude Sonnet 4.5 Anthropic Analyzing the Original Publication and Writing the Text The neural network studies the original material and generates a coherent text

1. Analyzing the Original Publication and Writing the Text

The neural network studies the original material and generates a coherent text

Claude Sonnet 4.5 Anthropic
2.
Gemini 3 Pro Preview Google DeepMind step.translate-en.title

2. step.translate-en.title

Gemini 3 Pro Preview Google DeepMind
3.
Gemini 2.5 Flash Google DeepMind Text Review and Editing Correction of errors, inaccuracies, and ambiguous phrasing

3. Text Review and Editing

Correction of errors, inaccuracies, and ambiguous phrasing

Gemini 2.5 Flash Google DeepMind
4.
DeepSeek-V3.2 DeepSeek Preparing the Illustration Description Generating a textual prompt for the visual model

4. Preparing the Illustration Description

Generating a textual prompt for the visual model

DeepSeek-V3.2 DeepSeek
5.
FLUX.2 Pro Black Forest Labs Creating the Illustration Generating an image based on the prepared prompt

5. Creating the Illustration

Generating an image based on the prepared prompt

FLUX.2 Pro Black Forest Labs

Related Publications

You May Also Like

Explore Other Events

Events are only part of the bigger picture. These materials help you see more broadly: the context, the consequences, and the ideas behind the news.

Want to dive deeper into the world
of neuro-creativity?

Be the first to learn about new books, articles, and AI experiments
on our Telegram channel!

Subscribe