Stop re-explaining
yourself to AI.

Switchy is one Space where your team, Claude, GPT-5, Gemini, and your MCP tools share memory. Decisions made today show up everywhere tomorrow — for you, your teammates, and every model you use.

Start free, no card
  • Free forever for solo
  • 350+ models
  • 1,000+ MCP tools
  • Shared team memory
  • Bring your own keys
  • Add teammates whenever

Or browse every model and every MCP integration first — pricing, benchmarks, and setup, no signup needed.

#product-launch
Message — @mention a teammate, model or tool…
Send
What Switchy does

The chat app where everything lives in one place.

Built around three pillars, mirrored in every surface: one thread for every model, every tool, every decision.

Every AI

One thread, every model

Claude, GPT-5, Gemini, Llama, Mistral — each mentionable inline. Pick the best mind for each turn. No tab-switching, no context loss.

Every tool

MCP tools as teammates

@github, @linear, @figma, @notion — any MCP server shows up in the @ picker. Switchy pulls context, runs the tool, stitches results into chat.

Every memory

Learns what you decide

Switchy watches for decisions, facts, and patterns — saves them as memories, then surfaces them again whenever another thread needs them.

The memory layer

A shared brain for your team, every AI, every tool.

Decisions, facts, and preferences become memory nodes. They travel with you across models, tools, and teammates — so the next conversation picks up where the last one stopped.

Decisions — with who, when, and the reasoning
Preferences — how your team likes things done
Facts — about people, products, and processes
Context — every model and tool sees the same truth
Works with every model — for the whole team

350+ models on one bill.

Switchy speaks to every provider. One usage budget across Claude, GPT-5, Gemini, Llama, Mistral, Grok, DeepSeek, Qwen — plus anything we add tomorrow.

Inside a Space, every teammate @-mentions the same models with the same shared context — pooled credits, pooled memory, no per-seat duplicates.

Claude
Opus · Sonnet
GPT-5
OpenAI
Gemini
2.5 Pro
Llama
Meta
Mistral
Large
Grok
xAI
DeepSeek
V3
Qwen
Max
Claude
Opus · Sonnet
GPT-5
OpenAI
Gemini
2.5 Pro
Llama
Meta
Mistral
Large
Grok
xAI
DeepSeek
V3
Qwen
Max
Transparent per-token pricingBring-your-own-key supportedFallback + load-balancing built inNew models live within 48h
Solo → team, zero friction

Start alone. Invite when you're ready.

Your solo workspace is already a team-capable Space. Add a teammate with a single email — they land in the same thread, sharing the same memory, signed in via their work email. Nothing to migrate.

Bring a brain, get a brain.

Invite your first teammate this week — both of you get +500 credits. About a month of average use, free.

  • No per-seat gotchas — pooled credits, pooled memory
  • Solo mode stays free forever
  • Team mode is a toggle, not a migration
  • Enterprise SSO + DPA when you need it
Start free and invite a teammate
Solo
Your personal workspace, free. 1 Space, 100 credits/mo, 1 MCP.
With one teammate
Both of you get +500 credits — about a month free. Upgrade to Team ($49/mo + $12/seat) when you outgrow it.
Same account · Same memory · Same @mentions
Security-first

Private by default, shared by choice

Every memory has a visibility — private, Space, or org. Your half-formed notes stay yours. Team members only ever see what you mark shareable.

  • SOC 2 — in progress
  • Data encrypted at rest + in transit
  • DPA available for every plan
  • No training on your data
Security details
Starter · free · solo forever

Pricing that scales with your team

Four tiers. Starter is free forever. Pro ($19/mo) unlocks unlimited Spaces and MCP. Team ($49/mo + $12/seat) shares memory across seats. Enterprise wraps it all with SSO + SOC 2 (in progress).

StarterPro $19Team $49 + $12/seatEnterprise
See all plans
Ready when you are

Your team. Every AI. One memory.

Start solo free — no card, no migration later when you invite teammates. Unlimited conversations. Every model. Every MCP tool.