polished handbook

This commit is contained in:
Ramon Perez 2025-08-11 14:50:04 +10:00
parent 72588db776
commit be659111a9
40 changed files with 591 additions and 534 deletions

Binary file not shown.

After

Width:  |  Height:  |  Size: 552 KiB

View File

@ -3,8 +3,6 @@ title: "ESOP"
description: "Employee Stock Options" description: "Employee Stock Options"
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Shares and Equity ## Shares and Equity

View File

@ -4,10 +4,6 @@ description: "Payment and benefits schemes."
asIndexPage: true asIndexPage: true
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Compensation
### Pay ### Pay

View File

@ -3,9 +3,9 @@ title: How We Communicate
description: Building a global AI platform requires clear, open communication description: Building a global AI platform requires clear, open communication
--- ---
# How We Communicate We're a fully-remote team building open superintelligence across continents
and cultures. Clear communication isn't just nice to have—it's how we ship fast,
We're a fully-remote team building open superintelligence across continents and cultures. Clear communication isn't just nice to have—it's how we ship fast, stay aligned, and build trust with millions of users. stay aligned, and build trust with millions of users.
## Core Communication Principles ## Core Communication Principles

View File

@ -1,3 +0,0 @@
---
title: Analytics
---

View File

@ -1,18 +0,0 @@
---
title: "Handbook"
description: "Jan Team Handbook"
---
## Github
Building in public is in our organizational DNA.
We open-source our code, share our learnings, and contribute back to the community. Sometimes it may feel like we're giving away our secret sauce, but we believe that the benefits of transparency and collaboration far outweigh the costs.
### Build in Public
- [Daily updates](https://discord.gg/AxypHJRQxd)
- [Project management](https://github.com/orgs/menloresearch/projects/5/views/25)
- [Open weights, open data](https://huggingface.co/homebrewltd)
- [Live training runs](https://discord.gg/BmA5DbxeEb)
- [Monthly community calls](https://discord.gg/J4aQxRUddD?event=1337239895755853846)

View File

@ -1,3 +0,0 @@
---
title: Infra
---

View File

@ -1,5 +0,0 @@
---
title: "Handbook"
description: "Jan Team Handbook"
---

View File

@ -1,10 +1,9 @@
--- ---
title: "Handbook" title: "Sprints"
description: "Jan Team Handbook" description: "Jan Team Handbook"
--- ---
## Sprints
[Jan sprints](https://github.com/orgs/menloresearch/projects/5/views/25) are every 2 weeks. [Jan sprints](https://github.com/orgs/menloresearch/projects/5/views/25) are every 2 weeks.
Releases are QA'd prior to release. We never release on Fridays, unless ya'll wanna come back on Saturday. Releases are QA'd prior to release. We never release on Fridays, unless ya'll
wanna come back on Saturday.

View File

@ -1,10 +1,8 @@
--- ---
title: "Handbook" title: "Internal Tools"
description: "Jan Team Handbook" description: "Jan Team Handbook"
--- ---
## Internal Tools
We like to make AI do the boring stuff. We also automate and dogfood our own tools as much as possible. We like to make AI do the boring stuff. We also automate and dogfood our own tools as much as possible.
### Tools ### Tools

View File

@ -3,11 +3,6 @@ title: "One on Ones"
description: "One on Ones." description: "One on Ones."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## 1-on-1s
We don't schedule recurring 1-on-1s, because in our experience, engineers and researchers [hate it](https://www.reddit.com/r/cscareerquestions/comments/14rkbwv/does_anybody_else_hate_their_11_meetings/). We believe it's better to address issues and provide feedback as they arise. We don't schedule recurring 1-on-1s, because in our experience, engineers and researchers [hate it](https://www.reddit.com/r/cscareerquestions/comments/14rkbwv/does_anybody_else_hate_their_11_meetings/). We believe it's better to address issues and provide feedback as they arise.
If you need to discuss something with us, simply book a time on the founder or team lead's calendar. **We're always available to chat.** If you need to discuss something with us, simply book a time on the founder or team lead's calendar. **We're always available to chat.**

View File

@ -3,11 +3,6 @@ title: "Interviewing"
description: "How to interview and hire outliers." description: "How to interview and hire outliers."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Interviewing
Our talent pipeline is an organizational moat. This guide walks you through our interviewing process. Our talent pipeline is an organizational moat. This guide walks you through our interviewing process.
### Process ### Process

View File

@ -3,11 +3,6 @@ title: "Time Off"
description: "Vacation and unplanned leave policy." description: "Vacation and unplanned leave policy."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Time Off
We offer 14 base vacation days, unlimited sick leave, and additional public holidays based on your respective region. We offer 14 base vacation days, unlimited sick leave, and additional public holidays based on your respective region.
### Planning Leave ### Planning Leave

View File

@ -3,11 +3,6 @@ title: "One on Ones"
description: "One on Ones." description: "One on Ones."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Career Progression
### Promotions ### Promotions
We don't believe in titles. We believe in fair compensation and fast progression. We don't believe in titles. We believe in fair compensation and fast progression.

View File

@ -3,9 +3,6 @@ title: "Retreats"
description: "Annual Team Retreat." description: "Annual Team Retreat."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Retreats ## Retreats
We host an annual team retreat to bring the team together, bond, and plan for the year ahead. We host an annual team retreat to bring the team together, bond, and plan for the year ahead.

View File

@ -3,11 +3,6 @@ title: "Side Gigs"
description: "Best practices around external engagements." description: "Best practices around external engagements."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Side Gigs
We have a permissive policy regarding side gigs, similar to [Posthog's](https://posthog.com/handbook/people/side-gigs). We have a permissive policy regarding side gigs, similar to [Posthog's](https://posthog.com/handbook/people/side-gigs).
However, our main ask is for Jan to be your "main gig", as you will be taking a crucial role in success/failure of Jan. Manage your time in a way that is fair to the company and your colleagues. However, our main ask is for Jan to be your "main gig", as you will be taking a crucial role in success/failure of Jan. Manage your time in a way that is fair to the company and your colleagues.

View File

@ -3,10 +3,6 @@ title: "Spending"
description: "Spending and reimbursement policy." description: "Spending and reimbursement policy."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Spending
We are incredibly frugal and intend to be around in 50 years. However, we avoid being "penny-wise & pound-foolish" and spend when it matters. We are incredibly frugal and intend to be around in 50 years. However, we avoid being "penny-wise & pound-foolish" and spend when it matters.

View File

@ -3,10 +3,6 @@ title: "Travel"
description: "Travel policy." description: "Travel policy."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Travel
We encourage team members to represent Jan at conferences and travel for in-person alignment We encourage team members to represent Jan at conferences and travel for in-person alignment

View File

@ -1,6 +1,6 @@
--- ---
title: "Jan Team Handbook" title: "Jan Team Handbook"
description: "Jan is building an open superintelligence platform that runs locally, privately, and offline." description: "Building superintelligence that you can own and run anywhere."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */}; {/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
@ -15,34 +15,68 @@ import Steps from '@/components/Steps.astro';
This handbook explains how [Jan](https://jan.ai) works, and is public. This handbook explains how [Jan](https://jan.ai) works, and is public.
Jan's Handbook is a [living document](https://en.wikipedia.org/wiki/Living_document), as Jan is constantly evolving. We're building superintelligence that you can self-host and use locally. Not as a limitation, but as a feature. Your AI should work wherever you need it - on your laptop during a flight, on your company's servers for compliance, or in the cloud for scale.
Jan's Handbook is a [living document](https://en.wikipedia.org/wiki/Living_document), constantly evolving as we build the future of AI ownership.
## 1. Chapter 1: Why does Jan exist? ## 1. Chapter 1: Why does Jan exist?
- [What problem are we solving?](/handbook/why/problem) - [What problem are we solving?](/handbook/why/problem) - Why AI needs to be ownable, not just rentable
- [Who we are building for](/handbook/why/icp) - [Who we are building for](/handbook/why/icp) - People who want flexibility and control
- [Our vision for open superintelligence](/handbook/why/ownership) - [Our vision for open superintelligence](/handbook/why/ownership) - An ecosystem you can own
## 2. Chapter 2: How we make money ## 2. Chapter 2: How we make money
- [How we make money](/handbook/money/how) - [How we make money](/handbook/money/how) - Open core + optional services
- [What is our moat](/handbook/money/moat) - [What is our moat](/handbook/money/moat) - Community, trust, and aligned incentives
## 3. Chapter 3: Who we hire ## 3. Chapter 3: Who we hire
- [The Fast and the Curious](/handbook/who/curiosity) - [The Fast and the Curious](/handbook/who/curiosity) - Ship today, learn for tomorrow
- [Underdogs Welcome](/handbook/who/underdogs) - [Underdogs Welcome](/handbook/who/underdogs) - Talent is everywhere, opportunity isn't
- [dy/dx > y-intercept](/handbook/who/dydx) - [dy/dx > y-intercept](/handbook/who/dydx) - Growth rate beats starting position
- [Obsession](/handbook/who/obsession) - We seek those who can't not build this
- [Team, not family](/handbook/who/team) - Pro sports team approach
## 4. Chapter 4: How we work ## 4. Chapter 4: How we work
- [Remote-first](/handbook/how/remote) - [Remote-first](/handbook/how/remote) - Global team, local impact
- [Open-source](/handbook/how/open-source) - [Open-source](/handbook/how/open-source) - Build in public
- [How we coordinate](/handbook/how/tools) - [How we coordinate](/handbook/how/tools) - Tools and processes
- [Sprints](/handbook/how/sprints) - [Sprints](/handbook/how/sprints) - Ship every week
- [How we communicate](/handbook/culture/communicate) - Clear, open, async-first
## 5. Chapter 5: Jan's Philosophy ## 5. Chapter 5: Jan's Philosophy
- [Open Superintelligence Platform](/handbook/philosophy/open-superintelligence) - [Open Superintelligence Platform](/handbook/philosophy/open-superintelligence) - Not one model, but an ecosystem
- [Humanity-aligned](/handbook/philosophy/humanity) - [Lateral Thinking with Proven Technology](/handbook/philosophy/lateral) - Innovation through integration
- [Perpetual Optimism is a Force Multiplier](/handbook/philosophy/optimism) - [Humanity-aligned](/handbook/philosophy/humanity) - Technology that unites
- [Perpetual Optimism](/handbook/philosophy/optimism) - Building the future we want to live in
## 6. Chapter 6: Team & Operations
- [Team Structure](/handbook/team) - How we're organized
- [Compensation](/handbook/comp) - Fair pay, meaningful equity
- [HR & Culture](/handbook/hr/onboarding) - Joining and growing with Jan
---
## Quick Links
- **For new team members**: Start with [Onboarding](/handbook/hr/onboarding)
- **For contributors**: Check out our [GitHub](https://github.com/janhq) and [Discord](https://discord.gg/FTk2MvZwJH)
- **For the curious**: Read about [our vision](/handbook/why/ownership)
## Our North Star
We're building superintelligence that:
- **Works anywhere**: From your laptop to your data center
- **Belongs to you**: Download it, own it, modify it
- **Scales infinitely**: One person or ten thousand, same platform
- **Improves constantly**: Community-driven development
This isn't just about making AI accessible. It's about ensuring the most transformative technology in human history can be owned by those who use it.
---
*"The future of AI isn't about choosing between local or cloud. It's about having both, and everything in between, working perfectly together."*

View File

@ -1,4 +0,0 @@
---
title: Interviewing at Jan
asIndexPage: true
---

View File

@ -1,31 +0,0 @@
---
title: "Offboarding from Jan"
description: "Offboarding Checklist."
asIndexPage: true
---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Offboarding
We are grateful for the time you spent with us. In many cases, you have moved on to start your own companies or to pursue your life's work. We'll be cheering you on from the sidelines and hope you remain part of the Jan community.
To ensure a seamless transition, we've outlined our offboarding process below.
### Notice Period
For voluntary departures past the probation period, we ask for 45 days notice both ways. This is to ensure a smooth transition for the team. We understand that this may not always be possible and will work with you to find a solution.
### Final Pay
Final pay & reimbursements will be processed at the end of the month with regular payroll. Unused vacations, equipment, and other considerations will also be netted in the final pay.
In terms of ESOP, we understand that most people are not in a position to exercise their options within a 90 day period. As a gesture of goodwill, we often grant a lengthy exercise period for good leavers. This is at the discretion of the board and is not guaranteed.
### Checklist
- Arrange to return equipment. If you wish to keep your equipment at a discounted rate, please inform HR.
- Assign all ongoing projects & Github issues to a team member.
- Optional: good leavers may request a recommendation letter from HR.
- Optional: request an exit interview to provide feedback on your experience at Jan.

View File

@ -4,11 +4,6 @@ description: "Onboarding Checklist for New Hires."
asIndexPage: true asIndexPage: true
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
## Onboarding
Welcome to Jan! We are excited to have you join our team. This guide walks you through the onboarding process. Welcome to Jan! We are excited to have you join our team. This guide walks you through the onboarding process.
> You will receive an onboarding invitation morning of your first day. Prior to the scheduled onboarding call, please send your Github & Discord handles to `hr@menlo.ai`. > You will receive an onboarding invitation morning of your first day. Prior to the scheduled onboarding call, please send your Github & Discord handles to `hr@menlo.ai`.

View File

@ -3,9 +3,9 @@ title: "How we make money"
description: "Building a sustainable business around open superintelligence" description: "Building a sustainable business around open superintelligence"
--- ---
# How we make money Jan is building a sustainable business that aligns our incentives with our users'
success. We believe the best way to build beneficial AI is to ensure our business
Jan is building a sustainable business that aligns our incentives with our users' success. We believe the best way to build beneficial AI is to ensure our business model reinforces our values: privacy, ownership, and universal access. model reinforces our values: privacy, ownership, and universal access.
## Our Business Model ## Our Business Model

View File

@ -3,9 +3,9 @@ title: "What is our moat?"
description: "Building defensibility through community, trust, and aligned incentives" description: "Building defensibility through community, trust, and aligned incentives"
--- ---
# What is our moat? Traditional moats rely on keeping secrets, locking in users, or maintaining
technical advantages. Jan's moat is different—it's built on openness, community,
Traditional moats rely on keeping secrets, locking in users, or maintaining technical advantages. Jan's moat is different—it's built on openness, community, and aligned values that are impossible to replicate with a closed approach. and aligned values that are impossible to replicate with a closed approach.
## Community as Competitive Advantage ## Community as Competitive Advantage

View File

@ -3,14 +3,8 @@ title: "Humanity-aligned"
description: "Jan exists because technology lets humanity work as one." description: "Jan exists because technology lets humanity work as one."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
# Humanity-aligned ![Our Blessed Kingdom](../../../../assets/tom_gauld.png)
<br />
<img src="/images/our-blessed-homeland.jpg" alt="Our Blessed Kingdom" width={800} height={450} />
> "The imagined community of the nation is always limited and sovereign, yet human connections and cooperation stretch far beyond borders." — Benedict Anderson > "The imagined community of the nation is always limited and sovereign, yet human connections and cooperation stretch far beyond borders." — Benedict Anderson

View File

@ -3,10 +3,6 @@ title: "Lateral Thinking with Proven Technology"
description: "Building the future with battle-tested tools" description: "Building the future with battle-tested tools"
--- ---
import YouTube from '@/components/YouTube.astro'
# Lateral Thinking with Proven Technology
> "Innovation is not about using the newest technology, but about using technology in new ways." — Adapted from Nintendo's philosophy > "Innovation is not about using the newest technology, but about using technology in new ways." — Adapted from Nintendo's philosophy
## Our Approach ## Our Approach
@ -18,7 +14,7 @@ At Jan, we don't chase the bleeding edge for its own sake. We take proven, battl
### Stability Over Novelty ### Stability Over Novelty
- **GGUF Format**: We didn't invent a new model format. We adopted the community standard that already works across platforms - **GGUF Format**: We didn't invent a new model format. We adopted the community standard that already works across platforms
- **llama.cpp**: Years of optimization by brilliant minds—why reinvent when we can build upon? - **llama.cpp**: Years of optimization by brilliant minds—why reinvent when we can build upon?
- **Electron + Web Tech**: Proven UI stack that works everywhere, letting us focus on AI innovation - **Tauri + Web Tech**: Proven UI stack that works everywhere, letting us focus on AI innovation
- **SQLite**: The world's most deployed database for local data—perfect for privacy-first architecture - **SQLite**: The world's most deployed database for local data—perfect for privacy-first architecture
### Accessibility Through Maturity ### Accessibility Through Maturity

View File

@ -1,213 +1,213 @@
--- ---
title: "Open Superintelligence Platform" title: "Open Superintelligence Platform"
description: "Building AI that augments humanity through openness, privacy, and universal access" description: "Building superintelligence as an ecosystem you can own and deploy anywhere"
--- ---
import YouTube from '@/components/YouTube.astro'
# Open Superintelligence Platform > "Superintelligence isn't one massive model. It's an ecosystem of specialized models, tools, and applications working together - and you should own all of it." — Jan Philosophy
> "The best way to predict the future is to invent it. The best way to invent it is to do it together." — Alan Kay (adapted)
## What is Open Superintelligence? ## What is Open Superintelligence?
Open superintelligence isn't just about making AI models public. It's a fundamental belief that the most powerful technologies in human history should be: Open superintelligence is AI that adapts to how you want to work, not the other way around. It's not about forcing you to choose between local or cloud, powerful or private, easy or extensible. It's about having it all, and owning it.
- **Owned by users**, not rented from corporations ### The Ecosystem Approach
- **Privacy-preserving**, not surveillance tools
- **Universally accessible**, not gated by geography or wealth
- **Community-driven**, not dictated by boardrooms
At Jan, we're building the platform where this vision becomes reality. We're not trying to build GPT-5. We're building something better: specialized models that excel at specific tasks, tools that extend their capabilities, and applications that tie it all together.
## Why Openness Matters ```
Models (specialized for tasks)
+ Tools (extend capabilities)
+ Applications (work everywhere)
= Superintelligence you own
```
### The Closed AI Problem ## Why Not One Giant Model?
Today's AI landscape is dominated by a few companies who: The "one model to rule them all" approach has fundamental flaws:
- Control what models you can use
- Decide what conversations are acceptable
- Store and analyze everything you type
- Can revoke access at any time
- Use your data to train their models
This isn't just inconvenient—it's dangerous. When a handful of corporations control humanity's access to intelligence augmentation, they control our collective future. - **Inefficient**: Using the same weights for poetry and mathematics
- **Inflexible**: Can't optimize for specific use cases
- **Expensive**: Massive compute for simple tasks
- **Monolithic**: Single point of failure and control
### The Open Alternative ### Our Approach: Specialized Excellence
Open superintelligence offers a different path: - **Jan-Search**: Knows how to find and synthesize information
- **Jan-Write**: Understands structure, tone, and creativity
- **Jan-Analyze**: Excels at reasoning and data interpretation
- **Jan-Code**: Optimized for programming tasks
**Transparency**: Know exactly how your AI works Each model does one thing brilliantly. Together, they form superintelligence.
- Inspect the code
- Verify the claims
- Understand the limitations
- Contribute improvements
**Ownership**: Your AI belongs to you ## The Three Pillars
- Download and run forever
- No subscription fees
- No terms of service changes
- No sudden deprecations
**Privacy**: Your thoughts remain yours ### 1. Models That Adapt
- Local execution
- No cloud dependencies
- Zero data collection
- Complete confidentiality
**Innovation**: Thousands of minds are better than hundreds Not just our models, but any model you need:
- Global collaboration - **Jan Models**: Purpose-built for specific tasks
- Diverse perspectives - **Community Models**: Any GGUF from Hugging Face
- Rapid iteration - **Cloud Models**: Connect to OpenAI, Anthropic when needed
- Community-driven features - **Your Models**: Fine-tune and deploy your own
## The Platform Philosophy ### 2. Tools That Extend
Jan isn't just an app—it's a platform for open superintelligence: Models need capabilities beyond conversation:
- **Search**: Get answers, not just links
- **DeepResearch**: Autonomous multi-step investigation
- **BrowserUse**: Let AI interact with the web
- **MCP Protocol**: Connect any data source or API
### 1. Best Models Win ### 3. Applications That Scale
We don't play favorites. Whether it's: Same experience, infinite deployment options:
- Our own Jan v1 model - **Jan Desktop**: Your local AI workstation
- Community contributions - **Jan.ai**: Browser-based, no setup required
- Open-source breakthroughs - **Jan Mobile**: AI that follows you (coming soon)
- Specialized domain models - **Jan Server**: Self-hosted for teams (coming soon)
If it performs well on **Jan Exam**, it belongs on our platform. Quality is our only metric. ## Ownership Through Flexibility
### 2. Composable Intelligence True ownership means having choices:
Intelligence isn't monolithic. Our platform enables: ### Choose Your Deployment
- **Multiple models** working together - **Full Local**: Everything on your hardware, works offline
- **Specialized capabilities** (vision, voice, reasoning) - **Full Cloud**: We host it, you just use it
- **Tool integration** via MCP - **Hybrid**: Local for sensitive work, cloud for heavy compute
- **Custom workflows** for specific needs - **Self-Hosted**: Your servers, your control
Users compose their own superintelligence from best-in-class components. ### Choose Your Models
- Use our models when they're best
- Use community models when they fit
- Use cloud models when you need them
- Train your own when you want to
### 3. Progressive Enhancement ### Choose Your Scale
- **Personal**: Run on your laptop
Start simple, grow sophisticated: - **Team**: Deploy on your server
- **Day 1**: Chat with a local model - **Enterprise**: Scale across infrastructure
- **Week 1**: Add tools and attachments - **Global**: Distribute to the edge
- **Month 1**: Fine-tune for your needs
- **Year 1**: Deploy across your organization
The platform grows with you.
### 4. Permissionless Innovation
Build whatever you imagine:
- No API keys to request
- No approval processes
- No platform restrictions
- No revenue sharing
Your innovation, your rules.
## Building in the Open ## Building in the Open
Our development process embodies our values: We don't just open source our code. We open source our entire development process:
### Public by Default ### Watch Us Train
- Open roadmaps Live at [train.jan.ai](https://train.jan.ai):
- Public discussions - See models training in real-time
- Transparent decisions - View our datasets and methods
- Community input - Learn from our failures
- Track actual progress
### Fail Fast, Learn Faster ### Test With Us
- Ship early and often Evaluate at [eval.jan.ai](https://eval.jan.ai):
- Learn from users - Compare model outputs
- Iterate based on feedback - Vote on what works
- Celebrate failed experiments - Suggest improvements
- Access all evaluation data
### Global Collaboration ### Build With Us
- Contributors from everywhere Everything on [GitHub](https://github.com/janhq):
- Diverse perspectives valued - Contribute features
- Local needs addressed - Report issues
- Universal solutions built - Fork and modify
- Join discussions
## The Network Effect of Openness ## The Network Effect of Openness
Open systems create compounding benefits: Open systems create compound benefits:
``` ### For Users
More Users → Better Feedback → Improved Models → - More models to choose from
More Contributors → Faster Innovation → More Users - More tools to extend capabilities
``` - More deployment options
- More control over your AI
Every user makes Jan better for everyone: ### For Developers
- Bug reports improve stability - Build on a stable platform
- Feature requests guide development - Access to growing ecosystem
- Model feedback enhances quality - No vendor lock-in fears
- Community support helps newcomers - Community support
## Our North Star: Jan Exam ### For Organizations
- Deploy how you need
- Customize for your use case
- Control your costs
- Own your infrastructure
Quality matters more than origin. **Jan Exam** ensures: ## Quality Without Compromise
- Objective performance metrics
- Transparent benchmarking
- Fair model comparison
- User-relevant testing
Whether a model comes from us, the community, or anywhere else—if it excels at helping users, it belongs on Jan. **Jan Exam** ensures excellence across the ecosystem:
- Objective benchmarks
- Real-world testing
- Community validation
- Transparent results
## The Future We're Building Whether it's our model or someone else's, if it performs well, it belongs on Jan.
Imagine a world where: ## The Path Forward
### AI Enhances Without Exploiting ### Today (Available Now)
- Your AI assistant knows you deeply but tells no one - Jan Desktop with local and cloud models
- Intelligence augmentation without surveillance capitalism - Basic tools via MCP
- Productivity without privacy compromise - Growing model ecosystem
- Active community
### Access Equals Opportunity ### Next 12 Months
- A student in rural India has the same AI as Silicon Valley - Jan v1 specialized models
- Language models speak every language, not just English - Advanced tools (search, research, browser)
- Costs approach zero as technology improves - Jan Server for teams
- Mobile applications
### Communities Shape Their Tools ### The Vision
- Local models for local needs - Models that understand your context
- Cultural values embedded in AI - Tools that act autonomously
- Bottom-up innovation, not top-down control - Applications that work everywhere
- An ecosystem owned by its users
### Intelligence Multiplies ## Why This Wins
- Every breakthrough benefits everyone
- Models improve through collaboration
- Knowledge compounds across communities
## Join the Movement ### Against Closed Platforms
- No vendor lock-in vs. total dependence
- Own forever vs. rent forever
- Infinite flexibility vs. their way only
- Community innovation vs. corporate roadmap
Open superintelligence isn't built by companies—it's built by communities. Every person who: ### Against DIY Open Source
- Complete ecosystem vs. fragmented tools
- Works out of box vs. endless configuration
- Unified experience vs. duct-tape integration
- Professional polish vs. research prototypes
- **Uses Jan** validates our mission ## Join the Revolution
- **Reports bugs** improves the platform
- **Shares feedback** shapes development
- **Contributes code** accelerates progress
- **Tells others** grows the movement
Is building the future with us. This isn't just about building better AI. It's about ensuring AI serves humanity, not the other way around.
When you use Jan, you're not just a user. You're part of a movement building:
- AI that works anywhere
- Intelligence you can own
- Tools that adapt to you
- A future without gatekeepers
## Our Promise ## Our Promise
We promise to: We promise to build superintelligence that:
1. **Stay Open**: Jan's core will always be open source 1. **Works Everywhere**: From laptop to data center to edge
2. **Respect Privacy**: Your data is yours alone 2. **Belongs to You**: Download it, modify it, own it forever
3. **Embrace Community**: Your voice matters in our decisions 3. **Stays Open**: Core will always be open source
4. **Pursue Quality**: Only the best for our users 4. **Keeps Improving**: Community-driven development
5. **Think Long-term**: Building for decades, not quarters 5. **Respects Choice**: Your deployment, your rules
## The Time is Now ## The Bottom Line
The race for AI dominance has begun. But while others race to lock users in, we're racing to set them free. Superintelligence shouldn't be controlled by a few companies. It shouldn't force you to choose between power and ownership. It shouldn't lock you into one way of working.
Open superintelligence isn't just possible—it's inevitable. The question is whether we build it together, or let it be built for us. With Jan, it doesn't.
At Jan, we've made our choice. We're building it together. You get superintelligence that adapts to how you work, runs where you need it, and belongs to you completely.
**Will you join us?** That's open superintelligence. That's what we're building.
--- ---
*"The future of intelligence is too important to leave to a few companies. It belongs to all of us."* *"The best AI platform is one where choosing local or cloud is like choosing between laptop or desktop - same experience, different form factor, your choice."*

View File

@ -5,10 +5,6 @@ description: "We believe in perpetual optimism."
import YouTube from '@/components/YouTube.astro' import YouTube from '@/components/YouTube.astro'
# Perpetual Optimism is a Force Multiplier
<br />
<YouTube id="o48X3_XQ9to?si=KDkCQ8MZe13sEc68" /> <YouTube id="o48X3_XQ9to?si=KDkCQ8MZe13sEc68" />
> "Success consists of going from failure to failure without loss of enthusiasm." — Winston Churchill > "Success consists of going from failure to failure without loss of enthusiasm." — Winston Churchill

View File

@ -4,14 +4,10 @@ description: "How we structure marketing to drive revenue and growth"
asIndexPage: true asIndexPage: true
--- ---
import Steps from '@/components/Steps.astro';
import YouTube from '@/components/YouTube.astro'; import YouTube from '@/components/YouTube.astro';
# Marketing
> "Marketing is about values." - Steve Jobs > "Marketing is about values." - Steve Jobs
<br />
<YouTube id="4mvHgLy_YV8?si=74cmdMmcH3gmpv0R" /> <YouTube id="4mvHgLy_YV8?si=74cmdMmcH3gmpv0R" />
Without a strategy, marketing is: Without a strategy, marketing is:

View File

@ -3,11 +3,11 @@ title: "Team Structure"
description: "How Jan's team is organized" description: "How Jan's team is organized"
--- ---
# Team Structure Jan is an open-source AI platform built by a distributed team focused on making AI
accessible, private, and locally-run.
Jan is an open-source AI platform built by a distributed team focused on making AI accessible, private, and locally-run. Jan is currently ~8 people working across different areas to build the open
superintelligence platform.
Jan is currently ~20 people working across different areas to build the open superintelligence platform.
## Jan Desktop & Platform ## Jan Desktop & Platform
@ -24,9 +24,16 @@ Focused on making the best models available to Jan users.
| Focus Area | Team Members | Description | | Focus Area | Team Members | Description |
|------------|--------------|-------------| |------------|--------------|-------------|
| **Inference Optimization** | Gau | Leading model quantization and evaluation for Jan's Local Model Hub - ensuring Jan supports the latest and best open source models | | **Model Hub** | Thien | Leading model quantization and evaluation for Jan's Local Model Hub - ensuring Jan supports the latest and best open source models |
| **Jan v1 Model** | Alan, Alex, Bách, Warren | Developing Jan's flagship model to be the default in Jan Desktop | | **Jan Models** | Alan, Alex, Bách, Warren | Developing Jan's flagship model to be the default in Jan Desktop |
| **Advanced Features** | Thinh, Norapat | Working with Ramon to spec and implement:<br/>• Jan Attachments<br/>• Jan DeepResearch<br/>• Jan Voice Mode | | **Advanced Features** | Thinh, Norapat | Working with Ramon to spec and implement:<br/>• Jan Attachments<br/>• Jan DeepResearch<br/>• Jan Voice Mode |
| **Engineering** | Akarshan, Louis, Dinh | Building the core of Jan |
| **Design** | Faisal | Making Jan look gooood, function well, and be accessible to everyone |
| **Product & DevRel** | Yuuki and Ramon | Thinking about the future of Jan and how to best communicate it to users |
| **Management** | Yuuki | Keeping people in check by threatening everyone with corporate speak |
| **Marketing** | Emre | Spreading the word about Jan and its capabilities |
| **Infra** | Minh and Hien| They make things run smoothly |
## Infrastructure & Operations ## Infrastructure & Operations

View File

@ -3,11 +3,6 @@ title: "The Fast and the Curious"
description: "We hire people who move quickly and never stop learning" description: "We hire people who move quickly and never stop learning"
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
# The Fast and the Curious
> "In the AI age, the ability to learn faster than the rate of change is the only sustainable advantage." — Adapted from Eric Ries > "In the AI age, the ability to learn faster than the rate of change is the only sustainable advantage." — Adapted from Eric Ries
We hire people who are both fast and curious—those who ship today while learning for tomorrow. We hire people who are both fast and curious—those who ship today while learning for tomorrow.

View File

@ -3,14 +3,10 @@ title: "dy/dx > y-intercept"
description: "The rate of a function is more important than its starting point." description: "The rate of a function is more important than its starting point."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
# dy/dx > y-intercept
> "A little bit of slope makes up for a lot of Y-intercept." — John Ousterhout > "A little bit of slope makes up for a lot of Y-intercept." — John Ousterhout
The rate of a function is more important than its starting point. We value growth rate over initial advantage. The rate of a function is more important than its starting point. We value growth
rate over initial advantage.
## Why slope matters ## Why slope matters

View File

@ -3,11 +3,6 @@ title: "Obsession"
description: "We seek obsession." description: "We seek obsession."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
# Obsession
> "If you want to build a ship, don't drum up the men to gather wood, divide the work, and give orders. Instead, teach them to yearn for the vast and endless sea." — Antoine de Saint-Exupéry > "If you want to build a ship, don't drum up the men to gather wood, divide the work, and give orders. Instead, teach them to yearn for the vast and endless sea." — Antoine de Saint-Exupéry
We don't hire for skills alone. We hire for obsession. We don't hire for skills alone. We hire for obsession.

View File

@ -3,11 +3,6 @@ title: "Jan is a Sports Team"
description: "Our approach is super simple: Retain. Build trust. Win together." description: "Our approach is super simple: Retain. Build trust. Win together."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
# Jan is a Pro Sports Team
> "First, you need to learn how to play. You need 2-3 years together to learn how to play with each other." — [Nikola Jokić](https://www.reddit.com/r/NBASpurs/comments/1cdscti/wise_words_from_jokic_first_you_need_to_learn_how/) > "First, you need to learn how to play. You need 2-3 years together to learn how to play with each other." — [Nikola Jokić](https://www.reddit.com/r/NBASpurs/comments/1cdscti/wise_words_from_jokic_first_you_need_to_learn_how/)
Jan is a sports team, not unlike a NBA team or a NFL team. We focus on putting the best players on the court, focusing on their 1% strengths, and building a team that can win together. Jan is a sports team, not unlike a NBA team or a NFL team. We focus on putting the best players on the court, focusing on their 1% strengths, and building a team that can win together.

View File

@ -3,11 +3,6 @@ title: "Underdogs"
description: "We hire from unconventional backgrounds." description: "We hire from unconventional backgrounds."
--- ---
{/* TODO: Replace with Astro-compatible icons: FaDiscord, FaMap */};
import Steps from '@/components/Steps.astro';
# Underdogs Welcome
> "I am, somehow, less interested in the weight and convolutions of Einstein's brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops." — Stephen Jay Gould > "I am, somehow, less interested in the weight and convolutions of Einstein's brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops." — Stephen Jay Gould
We hire from unconventional backgrounds. We hire from unconventional backgrounds.

View File

@ -1,5 +0,0 @@
---
title: "Handbook"
description: "Jan Team Handbook"
---

View File

@ -1,61 +1,112 @@
--- ---
title: "Who we are building for" title: "Who we are building for"
description: "Building for everyone who believes AI should be private, local, and owned" description: "Building for people who want to own their AI and choose how to run it"
--- ---
# Who we are building for Jan is built for people who want control over their AI experience. Our users span from individual developers to global enterprises, united by a desire to own their AI infrastructure and choose how they deploy it.
Jan is built for people who value privacy, ownership, and freedom in their AI interactions. Our users span from individual creators to enterprise teams, united by a common belief: AI should augment human capability without compromising human agency.
## Primary Users ## Primary Users
### Privacy-Conscious Individuals ### Developers Who Want Flexibility
People who understand that their thoughts, ideas, and conversations should remain private: Engineers and creators who need:
- Researchers working on sensitive topics - The freedom to run models locally or in the cloud
- Writers and creators protecting their intellectual property - Complete control over their AI stack
- Professionals handling confidential information - The ability to switch providers without losing their work
- Anyone who believes their AI conversations shouldn't train someone else's models - Integration options that fit their existing workflows
### Local-First Developers ### Individuals Who Value Ownership
Engineers and developers who need: People who understand that AI should be like any other software:
- Full control over their AI infrastructure - Download it once, use it forever
- The ability to run models offline and air-gapped - Your conversations and data belong to you
- Customizable, hackable AI tools - No vendor lock-in or subscription traps
- Integration with their existing workflows via our API - The freedom to modify and extend as needed
### Global Users with Limited Internet ### Teams That Need Options
Organizations that require:
- The flexibility to run on-premises or in the cloud
- The ability to scale from laptop to server cluster
- Control over costs and infrastructure
- Options that adapt to changing requirements
### Global Users Who Need Access
Millions of potential AI users who face: Millions of potential AI users who face:
- Unreliable or expensive internet connections - Expensive or unreliable cloud services
- Geographical restrictions on AI services - The need for AI that works offline
- The need for AI that works anywhere, anytime - Hardware limitations that require optimization
- Language and cultural barriers with centralized services - Different deployment needs for different situations
### Organizations Requiring Data Sovereignty ## Our North Star: Best Experience, Your Way
Companies and institutions that cannot use cloud AI due to:
- Regulatory compliance requirements
- Data residency laws
- Security policies prohibiting external data transmission
- The need for complete audit trails and control
## Our North Star: Quality for Everyone We use **Jan Exam** as our benchmark to ensure quality across all deployment options. Whether you're running locally on a laptop or scaled across a server farm, you get:
We use **Jan Exam** as our benchmark to ensure that whether you're using our models or any model from our hub, you're getting the best possible experience. We believe that: - The same powerful AI capabilities
- Consistent performance for your hardware
- The flexibility to change your setup anytime
- No compromise between local and cloud
- Quality shouldn't be gated behind subscriptions ## The Flexibility Spectrum
- The best models should be accessible to everyone
- Open models can and will outperform closed ones
- Community-driven development leads to better outcomes
## What We're NOT Building For Our users choose their own balance:
We're explicitly not building for: ### Full Local
- Surveillance or monitoring use cases - Everything runs on your hardware
- Applications that violate user privacy - Complete offline capability
- "Nice-to-have" enterprise vanity projects - Total data ownership
- Use cases that require sacrificing user control - Zero recurring costs
### Hybrid Approach
- Local for sensitive work
- Cloud for heavy compute
- Seamless switching between modes
- Optimize cost vs performance
### Full Cloud
- Jan.ai for zero setup
- Team collaboration features
- Managed infrastructure
- Scale without limits
## What Sets Our Users Apart
They understand that:
- **Ownership matters**: Your AI tools should be assets, not rentals
- **Flexibility is power**: Different tasks need different approaches
- **Control creates value**: The ability to customize and extend is crucial
- **Choice prevents lock-in**: Multiple options keep you free
## Real User Scenarios
### The Solo Developer
Runs Jan locally during development, uses cloud for production deployments. Owns their entire stack.
### The Research Team
Uses local models for sensitive data, cloud models for general research. Switches seamlessly based on needs.
### The Enterprise
Deploys Jan Server on-premises for compliance, uses Jan.ai for non-critical workloads. One platform, multiple deployment options.
### The Student
Runs lightweight models locally on their laptop, accesses more powerful models via cloud when needed. Learns AI without breaking the bank.
## Growing With Our Users
We build for users at every stage:
**Day 1**: Download Jan, start chatting locally
**Week 1**: Connect to cloud providers for more power
**Month 1**: Deploy your own server for team access
**Year 1**: Scale across infrastructure you control
## The Future User ## The Future User
Ultimately, we're building for the next billion AI users who haven't yet experienced what AI can do. They'll start with AI that respects their privacy, runs on their devices, and belongs to them from day one. We're building for the next billion AI users who will expect:
- Software they can own, not just rent
- The freedom to run AI anywhere
- Tools that adapt to their needs
- No artificial limitations
We're creating the platform where AI literacy begins with AI ownership. They'll start with Jan because it gives them choices, and they'll stay because those choices grow with them.
---
*"The best AI is the one that works where you need it, how you need it, when you need it."*

View File

@ -1,96 +1,180 @@
--- ---
title: "Our Vision for Open Superintelligence" title: "Our Vision for Open Superintelligence"
description: "Building AI that belongs to humanity, not corporations" description: "Building superintelligence that you can own, modify, and deploy however you want"
--- ---
# Our Vision for Open Superintelligence Jan believes the path to superintelligence isn't through one massive model controlled by a few companies. It's through an ecosystem of specialized models, tools, and applications that anyone can own and deploy.
Jan believes that the path to beneficial superintelligence is through openness, community ownership, and universal access. We're not just building another AI platform—we're creating the foundation for AI that truly belongs to its users.
## What is Open Superintelligence? ## What is Open Superintelligence?
Open superintelligence means: Open superintelligence is AI that:
- **Open Source**: Every line of code is public, auditable, and forkable - **Works Everywhere**: From your laptop to your data center to the edge
- **Open Weights**: Models that can be downloaded, modified, and owned - **Belongs to You**: Download it, modify it, deploy it - it's yours
- **Open Standards**: Protocols like MCP that anyone can implement - **Scales Infinitely**: Same AI whether you're one person or ten thousand
- **Open Participation**: A global community building together - **Evolves with Community**: Thousands of contributors, not hundreds of employees
## The Ecosystem Approach
Superintelligence isn't one thing - it's many things working together:
### Specialized Models
Instead of one model trying to do everything:
- **Jan-Search** excels at finding and synthesizing information
- **Jan-Write** understands creativity and structure
- **Jan-Analyze** specializes in reasoning and data
- **Jan-Code** focuses on programming tasks
Each model does one thing brilliantly, and they hand off tasks to each other.
### Extensible Tools
Models alone aren't enough. Our tools make them useful:
- Search the web and get answers, not links
- Automate browser tasks naturally
- Parse documents and understand context
- Connect to your data sources via MCP
### Flexible Deployment
One codebase, infinite configurations:
- **Local Mode**: Everything on your hardware
- **Server Mode**: Self-host for your team
- **Cloud Mode**: We host it for you
- **Hybrid Mode**: Mix and match as needed
## Why Ownership Matters ## Why Ownership Matters
When you use Jan, you own: When you own your AI infrastructure:
### Your Conversations ### You Control the Experience
- No company reads your chats - Customize models for your specific needs
- No AI trains on your prompts - Build workflows that match how you work
- Your data stays on your device - Integrate with your existing tools
- Export and move your data anytime - Scale based on your requirements
### Your Models ### You Keep What You Build
- Download once, use forever - Your prompts and conversations
- No subscription required - Your fine-tuned models
- Modify and fine-tune as needed - Your custom workflows
- Share with your community - Your accumulated knowledge
### Your Compute ### You Choose How to Run
- Run on your hardware - Offline when you need privacy
- No cloud dependency - Cloud when you need power
- Scale as you need - Local when you need speed
- Control your costs - Distributed when you need scale
## The Platform Approach ## Building in Public
We're building Jan as a platform where: We're not building this in secret. Everything is open:
- **Best Models Win**: Whether it's our Jan v1 model or community models, we showcase what performs best on Jan Exam ### Watch Us Train
- **Innovation Thrives**: Developers can build on top of Jan without permission See our models being trained in real-time at [train.jan.ai](https://train.jan.ai):
- **Standards Emerge**: Open protocols ensure no vendor lock-in - Live loss curves
- **Communities Flourish**: Users help users, models improve models - Training datasets
- Failed experiments
- Actual progress
## Community Ownership ### Test With Us
Help evaluate models at [eval.jan.ai](https://eval.jan.ai):
- Compare outputs side-by-side
- Vote on what actually works
- Suggest improvements
- See all the data
Jan is built by and for its community: ### Build With Us
Everything on [GitHub](https://github.com/janhq):
- Contribute code
- Report issues
- Suggest features
- Fork and modify
### Open Development ## The Path to Superintelligence
- Public roadmap and planning
- Community input on features
- Transparent decision making
- Open source contributions welcome
### Shared Success ### Today: Foundation (✅)
- Models improved by the community - Jan Desktop works with local and cloud models
- Features requested by users - Basic tools via MCP
- Quality benchmarked transparently - Community growing rapidly
- Success measured by adoption, not extraction
## The Path Forward ### Next 12 Months: Ecosystem
- Jan v1 models optimized for specific tasks
- Jan Server for self-hosted deployments
- Advanced tools (browser automation, deep research)
- Cross-device synchronization
We envision a future where: ### Future: True Superintelligence
- Models that understand your context
1. **Every device runs AI**: From phones to laptops to embedded systems - Tools that act on your behalf
2. **Every person owns AI**: Not rented, not surveilled, but truly owned - Applications that work everywhere
3. **Every community contributes**: Local models for local needs - An ecosystem that belongs to everyone
4. **Every innovation is shared**: Improvements benefit everyone
## Our Commitment ## Our Commitment
We commit to: We commit to building superintelligence that:
- **Stay Open**: Jan will always have an open-source version ### Stays Open
- **Stay Private**: We will never read your data or conversations - Core will always be open source
- **Stay Accessible**: Supporting diverse hardware and use cases - Models will always be downloadable
- **Stay Community-Driven**: Your needs drive our development - Standards will always be public
- Development will always be transparent
## Join Us ### Stays Flexible
- Run it anywhere you have compute
- Mix local and cloud as needed
- Scale up or down instantly
- Switch providers without friction
This vision requires a movement, not just a company. Whether you: ### Stays Yours
- No vendor lock-in
- No forced updates
- No subscription traps
- No data exploitation
- Use Jan for private conversations ## Why This Approach Wins
- Contribute code or models
- Share feedback and ideas
- Tell others about local AI
- Build on our platform
You're part of making open superintelligence a reality. ### Network Effects
Every user makes Jan better:
- Bug reports improve stability
- Feature requests guide development
- Model feedback enhances quality
- Community support helps everyone
The future of AI shouldn't be controlled by a few corporations. It should be owned by everyone who uses it. That's the future we're building at Jan. ### Compound Innovation
Open development accelerates progress:
- Best ideas come from anywhere
- Solutions shared instantly
- Problems solved collectively
- Innovation happens 24/7
### Aligned Incentives
We succeed when you succeed:
- Your productivity is our metric
- Your ownership is our philosophy
- Your freedom is our product
- Your success is our business model
## Join the Movement
This isn't just about building better AI. It's about ensuring the most powerful technology in human history belongs to humanity, not corporations.
Whether you:
- Use Jan for daily work
- Contribute code or ideas
- Share it with others
- Build on top of it
- Deploy it in your organization
You're part of building superintelligence that everyone can own.
## The Choice is Yours
In five years, AI will be everywhere. The question is:
**Will you rent intelligence from a monopoly, or will you own your own superintelligence?**
With Jan, you don't have to choose between powerful and private, between cloud and local, between easy and extensible.
You get it all. And it's yours.
---
*"The best superintelligence is the one you can run anywhere, modify for anything, and own forever."*

View File

@ -1,56 +1,106 @@
--- ---
title: "What problem are we solving?" title: "What problem are we solving?"
description: "Building open superintelligence that respects your privacy and freedom" description: "Building superintelligence that you can own and run anywhere"
--- ---
import YouTube from '@/components/YouTube.astro' Current AI platforms force unnecessary trade-offs:
# What problem are we solving? - **All or Nothing**: Either use their cloud or build everything yourself from scratch
- **Vendor Lock-in**: Your prompts, workflows, and knowledge trapped in one platform
Current AI platforms have fundamental problems: - **One-Size-Fits-All**: Same solution whether you're a student or an enterprise
- **Rent Forever**: Monthly subscriptions for software that should be yours
- **Centralized Control**: Your conversations, data, and AI capabilities are controlled by a few companies - **Limited Deployment**: Can't run where you need it - offline, on-premises, or at the edge
- **Privacy Violations**: Every prompt you type is stored, analyzed, and used to train models
- **Internet Dependency**: No connection means no AI - even for simple tasks
- **Subscription Lock-in**: Pay monthly forever or lose access to your own conversations
- **Limited Choice**: Stuck with one model provider's capabilities and biases
## Mission ## Mission
Jan is building an open superintelligence platform that runs locally, privately, and offline. We believe AI should be: Jan is building a complete AI ecosystem that adapts to how you want to work. We believe AI should be:
- **Yours to Own**: Download it once, use it forever - **Truly Owned**: Like any software, download it and it's yours forever
- **Private by Default**: Your thoughts stay on your device - **Infinitely Flexible**: Run locally, self-host, or use our cloud - same experience
- **Open and Flexible**: Use any model, from any provider, or none at all - **Professionally Complete**: Not just models, but the full stack of tools and applications
- **Universally Accessible**: Works on any hardware, anywhere in the world - **Universally Deployable**: From your laptop to your data center to the edge
## Approach ## The Full Stack Approach
We're building the full stack for open superintelligence: We're not building another chatbot. We're building superintelligence you can own:
- **Jan Desktop**: Local AI that runs on your computer today ### Models That Specialize
- **Jan Models**: Our own models optimized for privacy and performance - **Jan Models**: Purpose-built for specific tasks (search, analysis, writing)
- **Model Hub**: Access to the best open-source models, quantized and ready to use - **Community Models**: Any GGUF model from Hugging Face
- **Jan Exam**: Rigorous evaluation to ensure quality across all models - **Cloud Models**: Connect to OpenAI, Anthropic, or others when needed
- **Advanced Features**: - **Your Models**: Fine-tune and deploy your own
- Attachments for document understanding
- DeepResearch for comprehensive analysis ### Tools That Extend
- Voice Mode for natural conversations - **Search**: Get answers, not just links
- Tool calling via MCP for real-world capabilities - **DeepResearch**: Multi-step autonomous research
- **BrowserUse**: Let AI interact with the web for you
- **MCP Integration**: Connect to any data source or service
### Applications That Scale
- **Jan Desktop**: Your local AI workstation
- **Jan.ai**: Web access with no setup
- **Jan Mobile**: AI that follows you (coming soon)
- **Jan Server**: Your own private AI cloud (coming soon)
## Why This Matters ## Why This Matters
In 5 years, AI will be as essential as electricity. The question is: will you rent it from a monopoly, or will you own it? In 5 years, AI will be embedded in everything you do. The question isn't whether you'll use AI, but how:
We're ensuring the answer is ownership. Every feature we build, every model we support, every line of code we write is in service of this goal. **Option A**: Rent access from tech giants, accept their limitations, lose your data when you stop paying
**Option B**: Own your AI infrastructure, deploy it anywhere, keep everything you build
We're making Option B not just possible, but better than Option A.
## The Problem With Status Quo
### Cloud-Only Platforms
- Can't work offline
- Can't handle sensitive data
- Can't customize deeply
- Can't control costs
### DIY Open Source
- Requires deep technical knowledge
- Fragmented tools and models
- No unified experience
- Massive integration overhead
### Jan's Solution
- Download and run in minutes
- Complete ecosystem out of the box
- Scale from laptop to cluster
- Switch between local and cloud seamlessly
## Real Problems We Solve Today
### For Developers
"I want to use AI in my app but don't want vendor lock-in"
→ OpenAI-compatible API that runs anywhere
### For Teams
"We need AI but can't send data to the cloud"
→ Self-host Jan Server on your infrastructure
### For Individuals
"I want ChatGPT capabilities without the subscription"
→ Download Jan Desktop and own your AI
### For Enterprises
"We need to scale AI across different deployment scenarios"
→ One platform that works from edge to cloud
## Vision ## Vision
We envision a future where: We envision a future where:
- AI augments human intelligence without compromising privacy - AI runs wherever you have compute - laptop, phone, server, edge device
- Anyone can run superintelligence on their personal devices - You own your AI infrastructure like you own your other software
- Open models outperform closed ones through community collaboration - Models, tools, and applications work together seamlessly
- The next billion users experience AI on their own terms, not Silicon Valley's - The best AI isn't gated behind corporate APIs
Jan is the platform that makes this future possible. Jan is the platform that makes this future real, today.
---
*"The future of AI isn't about choosing between local or cloud. It's about having both, and everything in between, working perfectly together."*

View File

@ -1,8 +0,0 @@
---
title: "Handbook"
description: "Jan Team Handbook"
---
## How we got here
(Coming soon)