Build AI agents that fit your business
Launch knowledge-driven agents, mixed assistants, or full AI experiences from one platform. Connect your documents, tune the behavior, embed the widget, and go live fast.
A platform, not just a doc bot
Everything from first upload to post-launch tuning, in one place.
Choose how the AI behaves
Run a strict knowledge agent, a mixed assistant that prefers your docs, or a flexible assistant for broader conversations.
Reuse knowledge across agents
Keep a shared document library, link files to multiple agents, and promote good answers into fast knowledge overrides.
Customize and install the widget
Tune the greeting, launcher, size, color, and origin rules before copying a production-ready embed snippet.
Tune, measure, and scale
Track usage, review conversations, manage follow-ups, and get ranked suggestions to improve your agent over time.
See what you can build
Different teams can run different agent behaviors from the same workspace.
Customer Support Agent
Strictly answers from help center and support documentation.
Product Advisor
Uses your docs first, then keeps the conversation moving with flexible reasoning.
Onboarding Coach
Runs as a more flexible conversational assistant for guided onboarding and discovery.
Production-grade infrastructure
Security, reliability, and developer experience aren't afterthoughts — they're in the architecture.
Compliance & privacy
- GDPR compliant — data export, erasure, and portability
- SOC 2 Type I controls in place — audit-ready
- Zero critical vulnerabilities — OWASP Top 10 verified
- Append-only audit logs — immutable compliance trail
Reliability by design
- LLM circuit breaker with graceful cache fallback
- Redis Lua atomic credit locks — no double-charges
- Outbox pattern for guaranteed event delivery
- Per-tenant rate limiting with tier-based quotas
Developer-first
- OpenAPI 3.0 spec at /docs
- Structured JSON logs with trace IDs
- Centralized error logging with tenant-aware trace context
- Shadow DOM embed widget — zero CSS conflicts
- API keys for programmatic access (Pro & White Label)
Simple, transparent pricing
Pick the plan that matches your agent count, AI flexibility, and usage volume.
Starter
For solo operators and early validation.
- 1 agent
- 10,000 monthly credits
- Knowledge-only mode
- Website embed
- Basic analytics and a single-owner workspace
Growth
For teams and operators who need more agents, higher volume, and shared knowledge.
- Up to 5 agents
- 50,000 monthly credits
- Knowledge + assistant mode
- Shared docs across agents
- Agent duplication and embed customization
- Top questions, ranked actions, and webhook integrations
Pro
For serious production workloads and higher limits.
- Up to 20 agents
- 200,000 monthly credits
- All response modes
- Advanced analytics, audit-ready RBAC, and company-domain controls
- Higher rate limits and priority processing
- Priority support
- API keys for programmatic access
White Label
For agencies, resellers, and enterprise deployments.
- Branded widget experience
- Custom domains, brand controls, identity policy, and usage terms
- Premium support
- Tailored onboarding
- API keys for programmatic access
Compare plans
See exactly what's included at each tier before you decide.
| Feature | Starter | Growth | Pro | White Label |
|---|---|---|---|---|
| Usage | ||||
| Agents | 1 | Up to 5 | Up to 20 | Unlimited |
| Monthly credits | 10,000 | 50,000 | 200,000 | 200,000 + custom |
| Credit top-ups | ||||
| AI & modes | ||||
| Response modes | Knowledge only | Knowledge + assistant | All modes | All modes |
| AI model tier | Standard | Standard | Standard + Pro | Standard + Pro |
| Semantic cache | ||||
| RAG retrieval | ||||
| Knowledge overrides | ||||
| Shared document library | ||||
| Collaboration & controls | ||||
| RBAC (roles & permissions) | ||||
| Audit logs | ||||
| GDPR export & erasure | ||||
| Company domain controls | ||||
| Human handoff & follow-up queue | ||||
| Integrations | ||||
| Webhook integrations | ||||
| API keys | ||||
| Embed widget | ||||
| Branded widget (custom domain) | ||||
| Support | ||||
| Priority support | ||||
| Dedicated onboarding | ||||
| Feature | ||
|---|---|---|
| Usage | ||
| Agents | 1 | Up to 5 |
| Monthly credits | 10,000 | 50,000 |
| Credit top-ups | ||
| AI & modes | ||
| Response modes | Knowledge only | Knowledge + assistant |
| AI model tier | Standard | Standard |
| Semantic cache | ||
| RAG retrieval | ||
| Knowledge overrides | ||
| Shared document library | ||
| Collaboration & controls | ||
| RBAC (roles & permissions) | ||
| Audit logs | ||
| GDPR export & erasure | ||
| Company domain controls | ||
| Human handoff & follow-up queue | ||
| Integrations | ||
| Webhook integrations | ||
| API keys | ||
| Embed widget | ||
| Branded widget (custom domain) | ||
| Support | ||
| Priority support | ||
| Dedicated onboarding | ||
Frequently asked questions
Quick answers to the questions that come up most before signing up.
What is a credit?
Credits are the unit we use to measure AI usage. A typical response uses roughly 200–500 credits depending on length and mode — knowledge-only is the most efficient, full assistant mode uses the most. Credits reset monthly with your plan and can be topped up at any time.
What happens when I run out of credits?
Agent responses stop until you top up or your plan renews. No data is lost — your agents, documents, and settings stay intact. You can add more credits at any time from the billing page, and you'll receive an email notification when your balance is running low.
Can I change plans at any time?
Yes. Upgrades take effect immediately. Downgrades apply at the end of your current billing period.
Is there a free trial?
There is no time-limited trial, but you can start with Starter for $24/mo and cancel any time if it's not the right fit. No long-term commitment required.
Can I embed the agent on any website?
Yes. You get a Shadow DOM widget snippet that works on any site without CSS conflicts. You control which origins are allowed to load it.
How does RAG work?
You upload documents and they are chunked, embedded, and stored as vectors. When a visitor asks a question, the most relevant chunks are retrieved and injected into the AI prompt as context — so answers are grounded in your content, not hallucinated.
Is my data secure?
Every tenant is isolated with row-level security at the database level. Audit logs track all changes, and GDPR export and erasure APIs are available on Pro and above.
What is White Label?
White Label lets agencies and resellers deploy Eosphorus under their own brand — custom domain, branded widget, identity policy, and usage terms. Contact sales for pricing.
Can I export my data if I cancel?
Yes. You can export your conversation transcripts and visitor data at any time from the dashboard. GDPR-compliant data export and erasure APIs are available on Pro and above. Your data is never held hostage.
Do you train your models on my data?
No. Your documents and conversations are used solely to provide context for your specific chatbot. We do not use customer data to train foundation models.