Every AI model in one place

Every model.
One subscription.

Stop paying for Claude, ChatGPT and Gemini separately. Get them all here — with web search, a legal workspace with official court data, 370+ expert agents, vision, PDF, voice and image generation.

No card to start · cancel anytime

Claude Opus 4.7Claude Opus 4.8Claude Fable 5GPT-5.6 Luna ProGPT-5.6 Terra ProGPT-5.6 Sol Proo3Gemini 2.5 Pro61 curated
61
models available
+370
skills & agents
10
connected tools
1
subscription for everything
It's that simple

How it works

From zero to your first result in under a minute.

1
Step 1

Create your free account

No card. In seconds you're chatting with the best AIs.

2
Step 2

Pick a model or activate an agent

From the fastest to the smartest — there's a right model for every task.

3
Step 3

Chat, attach, generate images and talk

All in one place, real-time answers, automatic history.

Start free

No card to start · cancel anytime

Why Brains

Not just another AI chat

Features you won't find together anywhere else.

Auto Mode

The AI picks the best model for each question — code, writing or reasoning. You just ask.

Compare side by side

The same question across several models at once. See instantly which one answers best.

Memory that knows you

The AI remembers who you are between conversations: your profession, style and preferences.

Conversations everywhere

History synced between phone and computer — and you can install it as an app. Pick up where you left off.

Full legal workspace

A chat dedicated to law with guided shortcuts and case lookups in Brazil's official court database (CNJ).

Connected to the real world

Web search with cited sources + 10 tools: court cases, company records, financial indices, official gazettes, weather and more.

See it working

Three proofs, one from each part of the product. Nothing here is staged: the tool notices and the citation check are the ones the app shows.

Where to use it

One subscription, three places.

Editors

Extension published on the VS Code Marketplace and Open VSX — which covers Cursor, Windsurf and VSCodium.

VS CodeCursorWindsurfVSCodium

Terminal

CLI to use Brains without leaving the terminal, on the same plan.

npm i -g brains-cli

Browser

Chat, legal area and tools, nothing to install.

brainscopilot.com

There is no JetBrains or Eclipse extension yet. Docs

Legal area

Every citation is checked — or flagged

Each answer goes through a deterministic check: digests and statute articles are verified one by one against the official text. What does not exist gets a warning instead of slipping through.

3,392
digests from the STF, STJ, TST and TSE, with official wording
6,469
articles from the Constitution, Civil, Procedure, Criminal and Labour codes and more
0
invented citations that pass unflagged
Art. 319 of the CPC — checks out: “The initial petition shall state…”⚠️Art. 4444 of the CPC — does not exist in the official text. Confirm the number.
Initial petition
Guided form; you approve the structure before the document is drafted.
Case lookup
Real status from the Brazilian courts (DataJud/CNJ), inside the conversation.
Deadlines
Business-day counting with the year's holidays, plus e-mail and phone alerts.
Case file summary
Attach the PDF: parties, claims, status and what expires first.
See the legal area
Models

The best models in the world

From the global AI leaders. Switch models in the same chat, no extra account needed.

Claude Opus 4.7
Anthropic
1M context · recent release
Pro Max
Claude Opus 4.8
Anthropic
1M context · recent release
Pro Max
Claude Fable 5
Anthropic
1M context · recent release
Pro Max
GPT-5.6 Luna Pro
OpenAI
1M context · recent release
Pro Max
GPT-5.6 Terra Pro
OpenAI
1M context · recent release
Pro Max
GPT-5.6 Sol Pro
OpenAI
1M context · recent release
Pro Max
o3
OpenAI
Raciocínio avançado · 200k
Pro Max
Gemini 2.5 Pro
Google
1M context · recent release
Pro Max
Gemini 3.1 Pro
Google
1M context · recent release
Pro Max
Grok 4.3
xAI
1M context · recent release
Pro Max
Grok 4.5
xAI
Avançado · 500k
Pro Max
GPT-OSS 120B
Cerebras
Inferência ultrarrápida
Pro
Llama 3.3 70B
Groq / Meta
Versátil e rápido
Pro
Llama 3.1 8B
Groq / Meta
Ultra-rápido
Pro
Claude Haiku 4.5
Anthropic
Claude rápido e barato · 200k
Pro
Claude Sonnet 4.5
Anthropic
1M context · recent release
Pro
Claude Sonnet 5
Anthropic
1M context · recent release
Pro
GPT-4o mini
OpenAI
Econômico · 128k
Pro
GPT-5 mini
OpenAI
Rápido · 400k
Pro
GPT-5
OpenAI
Base sólida da OpenAI · 400k
Pro
GPT-5.1
OpenAI
Forte no dia a dia · 400k
Pro
GPT-5.6 Luna
OpenAI
1M context · recent release
Pro
GPT-5.6 Terra
OpenAI
1M context · recent release
Pro
GPT-5.6 Sol
OpenAI
1M context · recent release
Pro
o4-mini
OpenAI
Raciocínio econômico · 200k
Pro
GPT-5.3 Codex
OpenAI
Especialista em código · 400k
Pro
GPT-5.1 Codex Max
OpenAI
Código, mais capaz · 400k
Pro
GPT-5.1 Codex Mini
OpenAI
Código rápido e barato · 400k
Pro
Gemini 2.5 Flash Lite
Google
1M context · recent release
Pro
Gemini 2.5 Flash
Google
1M context · recent release
Pro
Gemini 3 Flash
Google
1M context · recent release
Pro
Gemini 3.6 Flash
Google
1M context · recent release
Pro
Gemini 3.5 Flash
Google
1M context · recent release
Pro
Gemini 3.1 Flash Lite
Google
1M context · recent release
Pro
DeepSeek V4 Flash
DeepSeek
1M context · recent release
Pro
DeepSeek V4 Pro
DeepSeek
1M context · recent release
Pro
DeepSeek V3.2
DeepSeek
Open-source forte · 164k
Pro
DeepSeek R1
DeepSeek
Raciocínio open-source · 164k
Pro
Qwen 3.7 Plus
Alibaba
1M context · recent release
Pro
Qwen 3.7 Max
Alibaba
Mais capaz da Qwen
Pro
Qwen3 Coder
Alibaba
1M context · recent release
Pro
Qwen3 Max
Alibaba
Topo Qwen 3
Pro
Kimi K3
Moonshot
1M context · recent release
Pro
Kimi K2.7 Code
Moonshot
Especialista em código · 262k
Pro
Kimi K2 Thinking
Moonshot
Raciocínio · 262k
Pro
Kimi K2.6
Moonshot
Geração recente · 262k
Pro
Mistral Medium 3.5
Mistral
Precisão europeia · 262k
Pro
Mistral Small 3.2
Mistral
Leve e barato · 131k
Pro
Mistral Large
Mistral
Flagship Mistral · 128k
Pro
Llama 4 Maverick
Meta
1M context · recent release
Pro
Llama 4 Scout
Meta
Contexto enorme · 10M
Pro
MiniMax M3
MiniMax
1M context · recent release
Pro
MiniMax M2
MiniMax
Rápido · 205k
Pro
GLM 5
Z.ai
GLM de topo · 203k
Pro
GLM 5.2
Z.ai
1M context · recent release
Pro
Nemotron 3 Ultra
NVIDIA
1M context · recent release
Pro
Nemotron 3 Super
NVIDIA
1M context · recent release
Pro
Nova Pro
Amazon
Multimodal Amazon · 300k
Pro
Nova Lite
Amazon
Barato · 300k
Pro
Command A
Cohere
Empresarial · 256k
Pro
GLM 4.7
Cerebras
Raciocínio rápido para código e agentes
Free
Specialist agents

370+ specialists ready to work

Activate by command on Pro, or let the AI detect and choose automatically on Pro Max.

Lawyer

Law

Analyzes contracts, spots unfair clauses and drafts legal documents with precise technical language.

ContractsFilingsCases

Senior Dev

Dev

Generates, reviews and explains code in any language. Finds bugs, suggests refactors and proposes architecture.

Code reviewDebugArchitecture

Data Analyst

Data

Reads spreadsheets and CSVs, uncovers hidden trends and builds executive reports in seconds.

ExcelSQLVisualization

Business Consultant

Business

Runs SWOT analysis, builds business plans, assesses pricing and spots market opportunities.

SWOTBusiness planMarket

Copywriter

Marketing

Writes copy that converts: headlines, emails, ads, posts and scripts in the right brand voice.

HeadlinesEmailAds

Personal Tutor

Study

Explains any concept from basics to advanced, creates exercises, grades answers and adapts the pace.

ExercisesReviewExams

Data Architect

Dev

Models schemas, writes optimized queries, proposes indexes and reviews migrations for any database.

SQLSchemaPerformance

Content Writer

Content

Produces articles, newsletters, video scripts and SEO-friendly captions with consistent tone.

SEOBlogNewsletter

Automation Specialist

Automation

Builds Python scripts, advanced formulas and workflows to eliminate repetitive tasks.

PythonZapierScripts

Automatic activation available on the Pro Max plan.

See all agents

Everything an AI chat needs

Modern features, in a clean and fast interface.

All the models

A closed catalog with selected Claude, GPT, Gemini, Grok, Kimi, DeepSeek, Llama and other models.

370+ skills & agents

Ready-made abilities that turn the AI into a specialist in your field — in English, Portuguese and Spanish.

Vision & PDF

Send images and PDFs (even scanned, with OCR). The AI reads, describes and analyzes.

Image generation

Describe it and get high-quality images generated right in the chat.

Two-way voice

Talk with the microphone, listen to the answers. Natural conversation with AI.

Private & secure

No prompt tracking. Export or delete your data anytime, with one click.

Connected to the real world

Web search with sources + 10 tools: court cases, company records, financial indices, official gazettes, weather…

Export to Word & PDF

Download any answer as .docx or .pdf, ready to deliver.

Share & install

Public links to share conversations, and install it as an app on your phone and computer.

Frequently asked questions

Which models are available?

The curated catalog has 60 models from leading labs and only shows options actually offered by Brains. Free uses GLM 4.7; Pro unlocks models such as Claude Sonnet 5, GPT-5.6 Luna and Gemini 3.5 Flash; Pro Max adds Claude Fable 5, Claude Opus 4.8, GPT-5.6 Luna Pro and Grok 4.5.

What are skills and agents?

Ready-made abilities that turn the AI into a specialist: lawyer, senior dev, data analyst, copywriter, tutor… 370+ in the library, working in English, Portuguese and Spanish. On Pro you trigger them by command; on Pro Max the AI detects the context and activates them itself.

Can the AI access the internet and real data?

Yes. Turn on web search for up-to-date answers with cited sources, and the connected tools for official data: court cases (CNJ/DataJud), company records, postal codes, financial indices, official gazettes, weather and more — available on every plan.

Is it useful for legal work?

It's one of our main focuses. The legal workspace has a dedicated chat, guided shortcuts (complaints, contracts, opinions, appeals, deadlines), case lookups in Brazil's official CNJ database and PDF case-file uploads. Results download as Word or PDF.

How does premium usage work?

More powerful models consume more of your premium usage; free models consume none. Premium usage renews every 5 hours with a weekly cap. When the window resets, you can still use the free models or buy a one-time top-up.

What are one-time top-ups?

One-off purchases (no subscription) that stay in your account and extend your premium usage when the 5-hour window runs out. Available on Pro and Pro Max. They never expire.

Do I need a card to start?

No. Free doesn't ask for a card. You only subscribe when you want more models and more usage.

Can I cancel anytime?

Yes. Monthly subscription, cancelled from the billing portal in one click. No lock-in.

Claude Fable 5, GPT-5.6 Luna Pro, Grok 4.5 and a curated catalog

Start now, no card

25 free messages every day. No expiration date. When you want more, just subscribe.