Pay for the conversations
Greenflash reads.
No per-seat creep. Unlimited products and team members on every plan.
Use Greenflash in the app, Claude, Copilot, Cursor, and any Skills-compatible agent.
| Line item | ·FreeGet started at no cost$0 / mo | RecommendedGrowthShip faster with full control$299 / mo | Compliance-readyEnterpriseBuilt for scale and complianceCustom |
|---|---|---|---|
| Monthly usage | 1M tokens | 10M tokens | Custom volume |
| Additional usage | Not included | $5 / 1M after 10M | Volume pricing |
| Workspace | Unlimited products + members | Unlimited products + members + roles | Unlimited + SSO |
| Custom analyses | 1 custom analysis | 3 custom analyses | 10 custom analyses |
| Agent skill | · | Analyze with Claude + Copilot | Analyze with Claude + Copilot |
| Workflow automation | 1 Slack channel | 3 Slack channels · 5 webhooks · Linear Integration | 10 Slack channels · 10 webhooks · Linear Integration |
| Admin controls | Basic | Spend limits | Audit trail · SOC 2 · HIPAA |
| Support | Community | Slack + email | Slack · team · SLA |
| Action | Try it free → | Choose Growth → | Talk to us → |
Usage notes
Token counting
Message content counts in full. System prompts and product context count at 25%. Tool calls, agentic steps, and reasoning traces are free.
Growth overages
10M tokens included, then $5 per 1M up to 250M / mo. Set a monthly spend cap and Greenflash pauses before exceeding it.
Enterprise
For higher volume, SSO, audit trails, HIPAA-ready workflows with BAA, dedicated support, and custom SLAs.
Your users are already showing you what to ship.Greenflash turns every conversation into product direction.
Your first two weeks
- Day 01See where users are getting stuck.
- Day 03Find the segments, accounts, and moments that need attention.
- Week 01Ship your first improvement with conversation-level evidence.
- Week 02Prioritize the roadmap from real usage, not anecdotes.
Free to start · no credit card · 1M tokens included
01
What is Greenflash?
Greenflash is product management for AI agents. We read every real user conversation, score it against your definition of quality, and tell you what to fix or ship next. Chat, voice, agentic workflows. All of it.
02
How is Greenflash different from AI observability or evals?
Observability watches your stack. Evals quiz your prompts. Greenflash reads the conversations your users are actually having in production, then ties what is broken to what to do about it.
03
What is the Product Quality Index (PQI)?
PQI is your quality philosophy, enforced. You weight four categories (User Satisfaction, Growth & Intent, Friction & Reliability, and Safety & Fairness) and every conversation, prompt, and product is ranked against that definition. Change the weights once. Greenflash applies them everywhere else.
04
What does Greenflash integrate with?
SDKs for Python and TypeScript/Node. A Claude Code plugin. A Skills CLI for any Skills-compatible agent. A public REST API for everything else. Pipe in business events from Stripe, PostHog, or your warehouse to tie conversations to actual revenue.
05
Does Greenflash work with voice agents?
Yes, and audio never touches our servers. We work from the transcript. Five analyzers run automatically on any call with per-turn timing: latency (TTFA), interruptions, silences, call-end, and prosody. Vapi, Retell, ElevenLabs, Bland, Synthflow, LiveKit, Pipecat, and OpenAI Realtime work out of the box.
06
How long does it take to set up Greenflash?
Roughly ten minutes. Faster if you live in Claude Code or any Skills-compatible agent: run `/greenflash-onboard` and we install the SDK, set up the client, and wire up conversation logging in your codebase ourselves.
07
Is Greenflash secure and compliant?
SOC 2 Type II. HIPAA-ready, with a BAA on Enterprise. Row-level tenant isolation, automatic PII redaction, and full audit logs on Enterprise. Your data stays yours.
08
Is there a free tier?
Yes. 1M tokens, no credit card. Enough to instrument a real product and see real findings before you commit to anything.