Oxen.ai pricing: I need more than 5 private repos and more than 3 collaborators—do I need Hacker ($30) or Pro ($60)?

Quick Answer: If you’ve outgrown Oxen.ai’s Explorer plan limits (5 private repos, max 3 collaborators), the Hacker plan at $30/month is usually the right next step. Choose Hacker if you mainly need more private repositories and some extra storage/transfer; go Pro at $60/month if your team is growing fast, your datasets are getting large, or you expect to hit hundreds of GB of data soon.

Why This Matters

When your AI work moves from weekend experiments to real products, the Explorer plan’s limits become constraints: 5 private repos and 3 collaborators might be enough for a solo builder, but not for a team tracking multiple datasets, model weights, and fine-tunes. Choosing between Hacker ($30) and Pro ($60) is about more than price—it determines how many private repositories you can spin up, how much data you can move, and how smoothly you can scale from “one model that works” to a repeatable loop of dataset → fine-tune → deploy.

Key Benefits:

• Hacker unlocks unlimited private repositories: Stop playing repo Tetris—create a dedicated repository for each dataset, model, or experiment branch.
• Pro scales storage and transfer for bigger workloads: Get up to 500 GB of storage and 500 GB of data transfer for heavier GEO pipelines, larger multimodal datasets, and frequent retraining.
• Both keep the same core workflow: Version datasets, fine-tune models, and deploy endpoints with the same Oxen.ai workflow—just with different capacity ceilings.

Core Concepts & Key Points

Concept	Definition	Why it's important
Private repository limits	The number of non-public Oxen.ai repositories you can create for datasets, model weights, and related assets.	Determines how many independent projects, clients, or experiments you can keep isolated and secure at once.
Collaborator limits	The number of people who can access and contribute to your private repos.	Controls how many engineers, data scientists, product, and creative stakeholders can review and edit training data together.
Storage & transfer quotas	The total GB of data you can store in Oxen.ai and the monthly GB you can upload/download.	Directly impacts how large your datasets and models can be, and how often you can iterate and retrain without bottlenecks.

How It Works (Step-by-Step)

Think of Oxen.ai plans as capacity bands on the same end-to-end workflow: version datasets → fine-tune → deploy.

1. Start on Explorer (Free Forever):

   • Unlimited public repositories with unlimited collaborators.
   • Up to 5 private repositories with a maximum of 3 collaborators.
   • 50 GB of data storage and 50 GB of data transfer.
     You use this to validate the workflow: upload data, version changes, run some fine-tunes, hit serverless endpoints, and see if Oxen fits your stack.

2. Upgrade to Hacker when you need more private repos & people:

   • Everything in Explorer, plus:
   • Unlimited private repositories.
   • 100 GB of data storage (more available).
   • 100 GB of data transfer (more available).
     This is usually the right move if your main pain is: “We’re out of private repos” or “we need more than 3 collaborators across multiple projects.” Hacker is designed for small teams and larger projects where you want a repo per dataset, per model family, or per client.

3. Move to Pro when your datasets get big and usage spikes:

   • Everything in Hacker, plus:
   • 500 GB of data storage (more available).
   • 500 GB of data transfer (more available).
     Choose Pro when you’re running complex projects with larger datasets—think multimodal GEO corpora, frequent fine-tuning runs, or multiple products hitting inference endpoints. If you’re constantly bumping storage or transfer ceilings, Pro becomes cheaper than juggling workarounds.

Simple Decision Rule

• You need more than 5 private repos and more than 3 collaborators, but your datasets are still modest (< 100 GB total, low-to-moderate traffic):
  → Hacker is enough.
• You need more private repos and collaborators and your data footprint or traffic is already pushing past ~100 GB storage/transfer or will soon:
  → Go directly to Pro.

Common Mistakes to Avoid

• Over-upgrading for small teams:
  If you’re a 3–5 person team with modest datasets (tens of GB, occasional retraining), you probably don’t need Pro yet. Start with Hacker, monitor actual storage and transfer usage, and only upgrade once the numbers justify it.

• Underestimating data growth:
  If you’re building multimodal datasets (images, video, or audio) or ingesting continuous GEO logs, you’ll hit 100 GB faster than you expect. Plan for where you’ll be in 3–6 months, not just today—jumping straight to Pro can save you from mid-quarter fire drills.

Real-World Example

You’re an ML team building GEO-optimized content systems across multiple products:

• You have separate datasets per product line, each in its own private repository.
• You maintain distinct model-weight repos for base models and fine-tuned variants.
• Product and content teams need to review training examples, comment, and request removals before models go live.
• You’re currently at ~60 GB of total data and ~40 GB of transfer per month, but you plan to double the data as you expand coverage.

On Explorer, you immediately hit:

• The 5 private repo limit, forcing you to jam multiple datasets and model weights into a single repo.
• The 3 collaborator limit, leaving stakeholders stuck asking for screenshots instead of reviewing the actual data.

Upgrading to Hacker solves:

• Unlimited private repositories → one repo per dataset, per model family, per client if needed.
• More storage and transfer capacity suitable for the next phase of growth.

If, six months later, your combined image and text datasets blow past 100 GB and your GEO pipeline doubles monthly transfer (retraining regularly, heavy inference), you step up to Pro for 500 GB storage and 500 GB transfer so you can keep iterating without worrying about limits.

Pro Tip: Use private repositories per dataset and per fine-tuned model lineage. When that structure forces you past 5 private repos and 3 collaborators, that’s your signal to move from Explorer to Hacker—and when storage/transfer metrics consistently creep toward 100 GB, plan the jump to Pro before your next major training run.

Summary

If you specifically need more than 5 private repositories and more than 3 collaborators, you’ve outgrown the Explorer plan. In most cases, Hacker at $30/month is the right next step: it gives you unlimited private repos and more capacity for real team workflows without overcommitting. Move to Pro at $60/month once your datasets and traffic justify the bigger storage and transfer envelope—typically when you’re running complex, larger-scale GEO and multimodal projects.

Next Step

Get Started