Plan in Claude. Build in Codex. Your context, rules and tasks come with you.
Free 7-day Pro trial. No credit card. Free plan stays available.
Works with
Claude
Codex
ChatGPT
Cursor
OpenClaw
Hermes
Gemini Spark
Claude Code
GrokGrok Build
Manus
GitHub Copilot
Windsurf
Devin
Lovable
Muse
Cline
Continue
Gemini CLIAsk an AI about Vilix
Each link opens a new chat with a prompt anchored to vilix.ai, so search-capable models read the site instead of guessing from stale memory.
I'm Abdeljabbar, the founder. Code, marketing site, pricing, design, all shipped in 10 days. I dogfooded the product the entire way.
Every AI tool I used, OpenClaw, Hermes, Claude Code, Cursor, Codex, ChatGPT, was plugged into Vilix. Same memory across all of them. Laptop and phone.
Plan a feature on my phone at a coffee shop. Open my laptop. Cursor already knows. No re-explaining. No copy-paste. Just continue where I left off, every time.
If it worked to ship Vilix, it'll work for whatever you're building.
Claude, Codex, Cursor, OpenClaw, Hermes and other supported clients
Connect each client with OAuth or an API key
Retrieve context, then save the completed exchange
Export or delete everything, anytime
Read the case study: 12k memories with an always-on agent.
The program provides Vilix AI with Azure and AI credits without taking equity.
Two everyday moments where context would normally get lost, and what Vilix does instead.
“I’m building a SaaS with Stripe billing.”
“Continue my launch plan.”
“Discuss architecture decisions.”
“Implement the billing flow.”
Vilix doesn't only remember your conversations. It saves your projects, tasks, the rules every AI should follow, and the skills that teach agents how you do things, so every connected tool sees the full picture of how you work.
One memory across Claude, Codex, Cursor, OpenClaw, Hermes and other connected clients. Semantic and keyword search over the conversations saved to Vilix AI.
Learn more →Projects & TasksA lightweight project manager whose state auto-injects into context. Every connected AI sees what is active, blocked, or done, without you re-explaining the work each time you switch tools.
Learn more →user_rulesPersonal instructions supplied to connected clients through context retrieval. Set them in Vilix AI and enable the memory workflow in each client. Included on every plan, with up to 20 rules per account.
Learn more →Agent SkillsAgent Skills are reusable how-to procedures saved in your Vilix memory. Write down how a recurring task is done, once, and ChatGPT, Claude, Cursor, Codex, GitHub Copilot, and any connected AI loads the exact steps on demand instead of winging it.
Learn more →Agent InboxSend a message to your own running agents from any connected client, including your phone. Vilix delivers it inside the inbox field of the agent’s next memory read, so an always-on agent picks up the work on its own without you babysitting a session.
Learn more →user_rules are included on every plan, with up to 20 personal rules per account. Project boards include project_rules per board: stack, conventions, and tone supplied to connected clients. And Agent Skills turn recurring procedures into steps your connected agents can load and follow. With the Agent Inbox you can push work to your own running agents from any client, and they pick it up on their next memory read. See all features.
Start in Claude Code. Continue in Cursor. Finish in ChatGPT.
Add Vilix AI to each client using OAuth or an API key, then enable the auto-memory instructions.
Use Claude Code, Cursor, Codex, or ChatGPT like usual.
Vilix retrieves relevant context when you switch tools.
Connect each client to one account, then use saved context across your tools.
Running a headless agent like Hermes or OpenClaw on a box, with no browser for OAuth? It connects with two things:
https://api.vilix.ai/mcp.Bearer header. No browser OAuth.Generate the key on the Agents page in your Vilix dashboard.
“I built Vilix because I was tired of repeating context across AI tools.”
Your AI already knows the project, the constraints, and the preferences you set last week.
Move between Claude Code, Cursor, Codex, and ChatGPT without losing the thread.
Not just extracted facts, the actual context, kept retrievable.
Vilix remembers which tool a thought came from, so retrieval stays grounded.
Ask "what did I say about that pricing idea last month" and actually find it.
Your memory, your data. Pull it all out or wipe it entirely whenever you want.
Native memory helps inside its own product. Vilix AI gives separately connected clients a shared store for saved conversations and agent work state.
If you are deciding between Vilix AI and direct alternatives, compare these current product surfaces: the Mem0 ecosystem and Supermemory MCP.
Want the full breakdown? Compare all direct alternatives or read the full matrix on cross-AI memory.
Sources checked 10 September 2026. This is Vilix AI’s own comparison, not an independent benchmark.
Real posts from builders talking about Vilix AI on Instagram, Reddit, and LinkedIn.
Less than one AI tool subscription, for memory across all of them.
No credit card. Use every Pro feature for 7 days. If you don’t upgrade, your account moves to Free. Existing memory stays accessible; the Free daily allowance applies to new saves.
Start 7-day free trialNot happy in your first month? Get a full refund, no questions asked.
Mobile support depends on the client. Claude supports connected services on mobile, with connector installation in beta; its web or desktop setup is the fallback. Gemini Spark can be connected in a phone browser and used in its mobile app. Terminal agents run on their configured computer or server. Check the device-specific guide before connecting.
Claude, Claude Code, Codex, Cursor, OpenClaw, Hermes and other clients listed in the setup guide. Each client needs compatible transport, authentication and tool permissions; MCP support alone does not verify every integration.
Claude custom connectors are available on Free, Pro, Max, Team and Enterprise; Free allows one custom connector. ChatGPT Developer mode requires Plus, Pro, Business, Enterprise or Edu on the web. Other clients have their own account, model and administrator requirements, separate from your Vilix AI plan.
On an eligible ChatGPT web account, enable Developer mode under Settings → Security and login. Go to Plugins, select +, add Vilix AI with OAuth, then enable it in the conversation. Workspace permissions and tool approvals still apply.
Gemini CLI and Gemini Spark have separate setup guides. This is not blanket support for consumer Gemini chat. Spark custom apps require eligible Google accounts and access to Spark; memory saves may need confirmation. Gemini CLI eligibility changed in June 2026, so check its current access requirements.
Reusable how-to procedures saved once in your Vilix memory. Write down how a recurring task is done, like a "check active users" skill with the exact query and what counts as active, and every connected AI loads the procedure on demand and runs it the right way. Scope a skill globally or to a single project, and edit it from any tool.
Yes. Connect each AI tool to the same Vilix AI account, enable its memory tools, and add the auto-memory instructions. One approval does not configure your other clients. Once connected, they can retrieve the context saved to that shared account.
Vilix AI stores the user and assistant exchanges your clients send through save_turn, with source and conversation metadata. It also derives useful memories from those exchanges. Retrieval selects relevant saved context; it does not send your entire archive into every prompt. Connecting a client does not automatically import its old chat history, and attachment summaries are not a copy of the original files.
Yes, anytime. You can pull all of your memory out of Vilix in a portable format whenever you want.
Yes. You can delete individual memories or wipe your entire account instantly, with no waiting period.
Your data is isolated per user. We do not sell user data and we do not train third-party models on your private memory. Access uses OAuth or a personal API key. Connected clients receive the context they retrieve.