StellarBase
StellarBase Platform Pricing

Pricing that scales
with your work.

Start solo, invite your team to explore, then scale into a self-hosted business — or take everything in-house. Same engine, every plan.

Looking for token / API pricing?

Two ways to run StellarBase — switch to bring the other plans forward.

Trial
€0 / 14 days

Try the whole platform — with your team.

Talk to sales
What's included
Invite your team — full collaboration during the trial
Advanced OCR (100+ languages, handwriting, forms, tables)
Multimodal search (text / image / audio / video)
AI agents (Researcher, Analyst, Writer, Reviewer…)
Multi-agent threads
1 Base, limited storage & capacity
Community support

Collaboration is included for the full trial so your whole team can explore. Solo plans start at Standard.

Standard
Soon

For solo power users who want real capacity.

Talk to sales
Everything in Trial (solo), plus
More storage, tokens and processing capacity
Multiple Bases
Full audit trail & inspectable processing
Community support

Standard is a single-user plan. Need your team back? That’s Pro.

Recommended
Pro
Soon

For teams handling sensitive data.

Talk to sales
Everything in Standard, plus
Team collaboration is back — multi-user, multi-agent, shared context
Automatic PII anonymization (Gate) — names, IDs, financial & medical fields protected before any AI sees them
Human-in-the-loop review of extractions
Slack / Teams integrationQ3 2026
Self-built workflowsQ3 2026
Bring your own model keys (BYOK)Q3 2026
Memory & learning from feedbackQ4 2026
Higher capacity across the board

Gate keeps sensitive fields out of every AI request, so you can use AI on regulated data without it leaving your control.

Business
Soon
Self-hosted platform · 5 capacity tiers

For mid-size and regulated teams that run the platform on their own infrastructure.

Talk to sales
Everything in Pro, plus
Self-host the core platform (Docker / K8s)
Five capacity tiers — pick your volume, scale up anytime
Bring your own Postgres / S3 / identity
Hybrid deployment (self-hosted + cloud services)
External triggers (n8n, Make, Zapier)Q3 2026
Custom processing pipeline (custom extractors)Q4 2026
Customer-managed keys (CMEK)
DORA / NIS2 ready
90-day audit logs
Enterprise
Custom
Fully self-hosted

For organizations whose policy requires data AND models in-house.

Talk to sales
Everything in Business, plus
Full self-hosted solution — the complete document-intelligence stack running on your own infrastructure, licensed to you
Air-gapped deployment
Bring your own model keys / Postgres / S3 / identity
Custom entity types, rules & dictionaries (Gate)Q4 2026
HITL approval workflows (Gate)Q3 2026
Dedicated support engineer + onboarding
Unlimited capacity
1 year+ audit logs

Running the models in-house is the Enterprise tier — it’s our core technology, licensed under a dedicated contract.

Full comparison

Every feature, across all plans

What's included where. Items marked with a quarter are on the roadmap — shown so you can see where each plan is heading; they aren't live yet.

✓ live · Q3 / Q4 2026 planned · — not included
Trial
14 days
Standard
Solo
Pro
Teams
Business
Self-host
Enterprise
In-house
Collaboration & users
Single user
Team collaboration
Trial only
Multi-user, multi-agent threads
Trial only
Shared context & mentions
Trial only
Collaborators
Team
1
Team
25–100
Unlimited
Document processing
Advanced OCR (100+ langs, handwriting, forms)
Complex tables & structures
Image understanding (charts, diagrams)
Human-in-the-loop review of extractions
Audio & video transcription
Soon
Soon
Soon
Soon
Soon
Custom processing pipeline
Q4 2026
Q4 2026
Full self-hosted solution
Anonymization (Gate)
Automatic PII anonymization
Detection (people, financial, medical, IDs)
Transparent proxy + tokenized handoff
Custom rules & dictionaries
Q4 2026
HITL approval workflows
Q3 2026
AI agents
Agent types (Researcher, Analyst, Writer…)
Goal-oriented loop
Zero-trust access chaining
Grounded in your data (citations)
Multi-agent collaboration & delegation
Internet & deep research
External actions (API, email, DB)
Memory & learning from feedback
Q4 2026
Q4 2026
Q4 2026
Bring your own model keys (BYOK)
Q3 2026
Q3 2026
Q3 2026
Workflows Q3 2026
Visual drag & drop editor
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Self-built workflows
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Q3 2026
API & YAML (code-first)
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Step controls (pause, stop, retry)
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Q3 2026
External triggers (n8n, Make, Zapier)
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Q3 2026
Search
Multimodal search (text/image/audio/video)
Agent-powered search (NL → synthesis + citations)
Search across relationships
Multi-language (40+)
Chat
Built-in chat (channels, threads)
REST API
WebSocket
Soon
Soon
Soon
Soon
Live process tracking
Slack / Teams integration
Q3 2026
Q3 2026
Q3 2026
Google Chat / Discord
Q4 2026
Q4 2026
Q4 2026
Security & compliance
AES-256 at rest, TLS 1.3 in transit
Zero-trust per-source authorization
RBAC (Viewer, Editor, Admin, Owner)
SSO (SAML 2.0 / OIDC)
Audit logs
7 days
30 days
30 days
90 days
1 year+
Data residency choice
Customer-managed keys (CMEK)
GDPR / EU Data Act
DORA / NIS2
ISO 27001
In progress
In progress
In progress
In progress
SOC 2 Type II
In progress
In progress
In progress
In progress
Deployment
Hosted by us (EU cloud)
Optional
Optional
Your identity (SAML, OIDC, Okta, Azure AD)
Your database (Postgres / RDS / Supabase)
Your storage (S3, GCS)
Self-host the core platform
Hybrid (self-hosted + cloud services)
Full self-hosted solution (models in-house)
Air-gapped deployment
Support
Community support
Priority email support
Dedicated support engineer
Onboarding assistance
Uptime SLA
99.5%
99.7%
99.7%

Common questions

Why does Standard not include team collaboration when the trial does?

The trial lets your whole team explore so you can see the value together. For ongoing teamwork, Pro brings collaboration back with multi-user threads and shared context. Standard is a focused single-user plan.

What is Gate / anonymization and why is it on Pro?

Gate detects and protects sensitive data — names, IDs, financial and medical fields — before any AI request, so the model never sees real values but your output is still correct. It’s the feature that lets regulated teams use AI on sensitive data at all. It’s available from Pro upward.

Can I run the models on my own infrastructure?

Self-hosting the OCR and models is the Enterprise tier, under a dedicated licence and contract, because that engine is our core technology. Business lets you self-host the platform itself while inference stays managed.

What’s marked Q3 / Q4 2026?

Features in active development with a target quarter. They’re shown so you can see where each plan is heading; they aren’t live yet.

How does Business capacity work?

Five tiers sharing the same features. Pick the volume you process and step up anytime — the fixed part stays flat, only capacity changes. The figures shown are preliminary estimates until the official launch.

Get Started

Ready to build your knowledge base?

Start your 14-day trial in 2 minutes. Scale into a team, a self-hosted business, or fully in-house when you’re ready.