The unit you pay with. $1 buys 1,000 credits, and Lyro draws them down as it resolves conversations — light models draw a few credits per resolution, stronger ones draw more.

Only when a conversation is resolved. There's no platform fee, no per-seat pricing, and no minimum — anything Lyro doesn't resolve costs nothing.

Can I use different models for different tasks?

Yes. Each agent — and each skill inside one — can run its own model, so you can put an inexpensive model on routine questions and a stronger one where accuracy matters most.

What happens when Lyro can't resolve something?

It hands the conversation to your team with the full history and context. Handoffs aren't resolutions, so they don't spend credits.

What is Enterprise vs Self-service?

Self-service is a credit wallet: sign up, top up, pay for what you use — no contract. Enterprise adds isolated infrastructure, your own model contracts and API keys, a guaranteed SLA and resolution rate, and a dedicated team on your account.

Is there a free trial?

Yes. Build your agent and test it against your past tickets before you spend anything — credits only go toward live resolutions.

Can I cap my monthly spend?

Yes. Lyro can only spend the credits you've topped up, and you can set alerts and hard limits so the bill never surprises you.

Pricing

Pay for what resolves, on the model you pick.

Each model burns credits at its own rate, so you can match price to ticket difficulty. Bringing your own model contract? See Enterprise.

Pay as you go

Self-service

Wallet-based. $1 = 1,000 credits. Each model burns at its own rate.

$200/mo

Choose a credit amount

1,0002,000,000

Get started

Top up any amount. Credits draw down only when Lyro resolves, at each model's own rate.

check_circlePay only for what you spend — no fixed fees
check_circleNative Claude Code & Codex integration
check_circleSwap Claude, GPT, Gemini per agent or per skill

Available models

Claude Haiku 4.5
Claude Opus 4.7
GPT-5 mini
Gemini Flash
Claude Sonnet 4.6
GPT-5
Gemini 2.5 Pro

Enterprise

Use your own contract (Anthropic, OpenAI, or Google). Lyro routes through your keys, billing you directly.

Talk to sales

Talk to sales for pricing on your usage profile.

check_circleIsolated infrastructure — private cloud in your preferred region
check_circleConnect your own model — any provider, your API keys
check_circleGuaranteed metrics — SLA and AI resolution rate

Enterprise exclusives

Private Cloud Instance
Connect Custom Model
Guaranteed SLA
Forward Deployed Engineer
Customer Success Manager
Training for your team
Dedicated Slack Channel
Guaranteed AI resolution rate
SSO

Savings calculator

See what switching really costs.

Set your volume, pick who you're on now. We'll show the gap.

How big is your support operation?

Tell us your team size, or enter your monthly volume directly.

People on your support team

1250

We'll estimate about 945 conversations a month for each person on your team → 4,725 / mo.

What are you using today?

We'll compare your numbers against this.

Intercom (Fin) bills a flat $0.99 per resolution, no matter how simple or complex the conversation.

Recommended for you

Pay as you go

Wallet-based. Pay only for what resolves — no seats, no platform fee.

$723/mo

Get started

Annual savings: $8,158
Annual savings
Time back for your team: 265hrs/mo
Time back for your team

4,725 conversations/mo · vs Intercom (Fin)

Talk to sales

What's included

Same platform. Different billing.

Everything below ships in both plans. Enterprise adds private infrastructure, your own model keys, guaranteed metrics, and a dedicated team.

Feature	Self-service	Enterprise

Private Cloud Instance — your data fully isolated, in your preferred region	cancel	check_circle
Connect Custom Model — any model, your own API keys	cancel	check_circle
Guaranteed SLA	cancel	check_circle
Dedicated Forward Deployed Engineer	cancel	check_circle
Dedicated Customer Success Manager	cancel	check_circle
Training for your team	cancel	check_circle
Dedicated Slack Channel	cancel	check_circle
Guaranteed AI resolution rate	cancel	check_circle
SSO	cancel	check_circle

Private Cloud Instance — your data fully isolated, in your preferred region
Self-service
cancel
Enterprise
check_circle
Connect Custom Model — any model, your own API keys
Self-service
cancel
Enterprise
check_circle
Guaranteed SLA
Self-service
cancel
Enterprise
check_circle
Dedicated Forward Deployed Engineer
Self-service
cancel
Enterprise
check_circle
Dedicated Customer Success Manager
Self-service
cancel
Enterprise
check_circle
Training for your team
Self-service
cancel
Enterprise
check_circle
Dedicated Slack Channel
Self-service
cancel
Enterprise
check_circle
Guaranteed AI resolution rate
Self-service
cancel
Enterprise
check_circle
SSO
Self-service
cancel
Enterprise
check_circle

AI Chatbot — autonomous, all channels
Self-service
check_circle
Enterprise
check_circle
LLM Providers — Anthropic · OpenAI · Google, per-agent picker
Self-service
check_circle
Enterprise
check_circle
RAG / Knowledge Base Q&A
Self-service
check_circle
Enterprise
check_circle
Agentic Actions (tools)
Self-service
check_circle
Enterprise
check_circle
AI Copilot for Human Agents + Suggested Replies
Self-service
check_circle
Enterprise
check_circle
AI Ticket / Intent Classification
Self-service
check_circle
Enterprise
check_circle
AI Conversation Summary
Self-service
check_circle
Enterprise
check_circle
AI-Predicted CSAT
Self-service
check_circle
Enterprise
check_circle
AI Translation — replies in visitor's language + two-way inbox translation
Self-service
check_circle
Enterprise
check_circle
No-Code AI Setup — crawl → agent → playground → deploy
Self-service
check_circle
Enterprise
check_circle
AI Voice Bot (phone)
Self-service
check_circle
Enterprise
check_circle
AI Quality Assurance — Sentry auditor + LLM-judge evals
Self-service
check_circle
Enterprise
check_circle
Training Data — Crawl · CSV · PDF · Confluence · Outline · articles
Self-service
check_circle
Enterprise
check_circle

Custom HTTP/GraphQL tools — chains, secrets, test runner
Self-service
check_circle
Enterprise
check_circle
Bring-your-own MCP servers
Self-service
check_circle
Enterprise
check_circle
First-party MCP server — manage Lyro from any MCP client
Self-service
check_circle
Enterprise
check_circle
1k+ integrations
Self-service
check_circle
Enterprise
check_circle
REST API + outbound webhooks
Self-service
check_circle
Enterprise
check_circle
Workspace secrets vault · format presets
Self-service
check_circle
Enterprise
check_circle

Multiple versioned agents per workspace
Self-service
check_circle
Enterprise
check_circle
Skills & procedures system
Self-service
check_circle
Enterprise
check_circle
Model picker + thinking level per agent
Self-service
check_circle
Enterprise
check_circle
Guardrails & prompt-injection defenses
Self-service
check_circle
Enterprise
check_circle
Playground
Self-service
check_circle
Enterprise
check_circle
Evals & datasets
Self-service
check_circle
Enterprise
check_circle
Sentry auto-QA + KB gap analyzer
Self-service
check_circle
Enterprise
check_circle
Operator onboarding copilot
Self-service
check_circle
Enterprise
check_circle
Live takeover & release-to-AI, two-mode escalation engine
Self-service
check_circle
Enterprise
check_circle
Distribution behavior config
Self-service
check_circle
Enterprise
check_circle
Widget Custom Branding
Self-service
check_circle
Enterprise
check_circle

FAQ

Pricing, without the asterisks.

Straight answers on credits, models, and what each plan actually costs.

The unit you pay with. $1 buys 1,000 credits, and Lyro draws them down as it resolves conversations — light models draw a few credits per resolution, stronger ones draw more.

Two support agents collaborating at a laptop

Go live in a day, not a quarter.

Connect your help desk, train Lyro on your past tickets, and watch it resolve before you commit. No consultants, no implementation project.

Get started Talk to sales