Introducing Radius: The AI workspace for teams

The next-generation AI assistant for your team

Context-aware chats, automated codebase diagnostics, and deep integration with your engineering and ticketing stack. Bring the power of state-of-the-art language models directly into your workspace.

Get Started
Hero background
Dashboard preview
Claude
Turso
OpenAI
Clerk
Github
Supabase
Nvidia
Stripe
Claude
Turso
OpenAI
Clerk
Github
Supabase
Nvidia
Stripe

Why Radius?

The core model workspace
that grows with your needs

Other assistants live in browser tabs, disconnected from your actual workspace tools. Radius brings AI directly into your developer terminal, documents, calendars, and tickets, keeping your operations fully integrated.

All-in-one assistant workspace

Chat with AI models, draft documents, run diagnostics, and trigger external automations from a single context-aware dashboard. No more hopping between tabs.

diagnostics
Olivia Hartley9:14

@Radius run diagnostics on client component dependencies and draft a fix

Radius9:14

Found build error in next.config.ts due to missing @next/mdx loader. Drafted fix in branch 'fix/mdx-loader' and ran verification build.

Olivia Hartley9:15

Love this — saved me an afternoon.

Collaborate across Slack, terminal, or Web

Whether you trigger a prompt via Slack DM, terminal, or our Web canvas, Radius keeps your context and generation history fully synchronized.

Frontier models

Claude logo
Claude 3.5 Sonnet

Deep codebase reasoning and dynamic code editing

OpenAI logo
GPT-4o

Fast conversational speed and image inputs

DeepSeek logo
DeepSeek R1

Long-context logical deductions and code tracing

Swap frontier models on the fly

Switch between advanced models depending on the complexity of your prompts. Choose speed, intelligence, or deep reasoning at any step.

Model workspace

From prompts to production,
no manual steps

Code diagnostics, repo monitoring, status reports, and integrations run automatically. Your team never has to track it manually.

Context-aware chats

Seamlessly reference files, folders, issues, and code directly inside chat messages. The model stays fully aligned with your project.

Dynamic code generation

Autocomplete, refactor, and generate clean scripts and features with live compilation feedback. Accelerate development with zero syntax struggles.

Automated workflows

Set up automated agents that monitor repositories, update tickets, and summarize drafts on schedules. Free up senior engineers.

Interactive artifact views

Iterate on UI mockups, documents, and markdown previews side-by-side inside the chat workspace. Design and review in real-time.

Granular permissions

Folder-level access control, role-based visibility, and full audit logs. Sensitive database credentials and customer records stay secure.

Direct integrations

Sync with GitHub, Slack, Jira, Notion, and production databases to trigger complex actions using natural language.

Agentic workspaces

Set up a goal. Watch it run.

Define tasks in plain English. Radius plans the execution steps, scans files, modifies code, and runs validations autonomously.

Autonomous execution

Create agentic routines that inspect codebases, execute terminal testing scripts, and summarize findings. Fully hands-off.

Integrated toolchains

Pass variables and outputs seamlessly between compiler systems, database migrations, and ticketing APIs. No manual glue.

Grounded reasoning

Model outputs are directly grounded in your repository files, workspace structures, and Slack channels—avoiding generic AI hallucinations.

Fix the next/mdx compilation build error in app/blog

mdx compilation plan

Read codebase files

Analyzed 12 files in client directory

Diagnose build errors

Found unresolved external module node:fs/promises

Modify page components

Converted index page to Server Component

Wrap config with next/mdx

Configured pageExtensions and MDX compiler

Run build validation

Turbopack build compiled in 2.4s

Publish change to git

Pushed branch fix/mdx-loader to origin

Radius Synthesis · Research Report

Reasoning

We compiled feedback, issue reports, and source files across your repository. The context suggests the main bottleneck is related to server-only modules imported in client files. We recommend refactoring component boundaries to use next/mdx loaders directly…

Sources
Google logoSlack logoLinear logoGitHub logo

Radius updated research file

Sourced from GitHub issues and Slack threads

Context synthesis

Deep context compilation,
not generic answers

Compile notes, chats, issues, and documents into comprehensive summaries grounded in real team data.

Cross-platform context sourcing

Connect to your engineering and operations stacks to automatically ground chats in real documentation and tickets.

Dynamic reports and summaries

Generate specifications, architecture overviews, and test reports with up-to-date repository insights.

Interactive editing

Iterate on drafted content inside the visual canvas workspace with real-time feedback loops.

Built for

Different roles. Same product.

Developer working on code

Deep codebase reasoning for

Developers & Engineers

Diagnose compilation errors, write integration tests, and build features directly in your repo context.

Designer collaborating in canvas

Interactive layout designs for

Designers & Creators

Iterate on mockups and user experiences side-by-side with real-time markdown previews.

Product manager working at a laptop

Dynamic research compiling for

Product & Operations

Draft specs, compile sprint reviews, and manage issue routing using grounded workspace data.

Pricing

Simple pricing,
built to scale with you

One flat per-user fee. No seat tiers, no add-on modules, no surprise integration costs.

Starter

An AI companion to boost your day-to-day productivity.

$75/mo
  • Up to 5 team members
  • Access to standard models (GPT-4o mini, Claude Haiku)
  • Basic file and document uploads
  • Slack integration
  • Standard response times

Growth

Collaborative AI features for active engineering and product teams.

$150/mo
  • Unlimited team members
  • Everything in Starter
  • Advanced models (Claude 3.5 Sonnet, GPT-4o, DeepSeek R1)
  • Deep codebase index & file context
  • Model Context Protocol (MCP) integrations
  • Priority chat support

Enterprise

Advanced compliance, custom models, and dedicated environments.

Let's talk
  • Everything in Growth
  • SOC 2 Type II, SSO/SAML, SCIM
  • Zero-data-training guarantee (private API nodes)
  • Dedicated model endpoints and custom prompt rules
  • Dedicated CSM and custom SLAs

Prices exclude VAT/GST. Volume discounts apply automatically for larger teams. All plans include unlimited integrations.

Security & Privacy

Enterprise-grade data safety

SOC 2 Type II compliant, SSO/SAML authorization, folder-level permissions, and zero-data-training guarantees to keep your proprietary codebase secure.

AICPA SOC certified
SSO / SAML supported

Integrations

Works where your team
already does

ask-radius
Hannah Reid14:02

How do I resolve the next/mdx compilation build error?

Radius14:02

You need to configure createMDX inside next.config.ts and register the loader for .mdx file extensions. Detailed steps are in docs/mdx-loader §2.1.

Hannah Reid14:03

@Radius draft the next.config.ts configuration wrapper for me

Slack agent

A collaborative assistant for every developer, right in Slack. Troubleshoot build logs, check PR statuses, and query project documentation from the chat window your team lives in.

Search repository files for all uses of node:fs/promises in client page components.

Radius

MCP interface

Connect Radius to Claude web, Claude Code, Cursor, or any MCP-compatible host. Every workspace action is available over MCP, so it slots into your automation workflows too.

Slack logo
Google logo
Microsoft logo
Notion logo
GitHub logo
GitLab logo
Jira logo
Linear logo
Vercel logo

Platform integrations

Sync with GitHub, GitLab, Slack, Jira, Notion, Linear, Google Workspace, Okta, and Vercel. Codebase context flows securely between your platforms.

FAQ

Everything you need
to know

Answers to the questions we hear most. Still curious? We're happy to chat.