CogniSpend

Pricing

Start free. Scale when you have the data to justify it.

No per-seat fees. No surprise usage charges. Pricing scales with the volume of AI events you track.

Free
$0/ mo

For individual engineers and small teams getting started.

Start free
  • 1 project
  • 10,000 AI events per month
  • 7-day data retention
  • SDK and proxy integrations
  • Manual cost entries
  • Basic cost dashboard
  • Community support
ProMost popular
$49/ mo

$39/mo billed annually

For engineering teams running AI in production.

Start Pro trial
  • Unlimited projects
  • 1,000,000 AI events per month
  • 90-day data retention
  • Full ROI measurement
  • Automated insights engine
  • Budget alerts and Slack notifications
  • Team management and RBAC
  • CSV import for historical data
  • Email support
Enterprise
Custom

For organizations that need unlimited scale and enterprise controls.

Talk to us
  • Unlimited events and retention
  • SSO / SAML
  • Custom data residency
  • Dedicated onboarding and support
  • SLA with uptime guarantee
  • Custom integrations
  • Audit log
  • Annual invoicing

All prices in USD. Annual pricing billed yearly. Event volume resets monthly. Pricing subject to change during beta.

Full feature comparison

Every plan. Every feature. No surprises.

Free

  • Projects1
  • AI events / month10,000
  • Data retention7 days
  • Team members1
  • SDK wrappers
  • Proxy mode
  • ROI measurement
  • Automated insights
  • Budget alerts
  • CSV import
  • Copilot / Cursor tracking
  • REST API access
  • SSO / SAML
  • Uptime SLA
  • SupportCommunity

Pro

  • ProjectsUnlimited
  • AI events / month1,000,000
  • Data retention90 days
  • Team membersUnlimited
  • SDK wrappers
  • Proxy mode
  • ROI measurement
  • Automated insights
  • Budget alerts
  • CSV import
  • Copilot / Cursor tracking
  • REST API access
  • SSO / SAML
  • Uptime SLA
  • SupportEmail

Enterprise

  • ProjectsUnlimited
  • AI events / monthUnlimited
  • Data retentionUnlimited
  • Team membersUnlimited
  • SDK wrappers
  • Proxy mode
  • ROI measurement
  • Automated insights
  • Budget alerts
  • CSV import
  • Copilot / Cursor tracking
  • REST API access
  • SSO / SAML
  • Uptime SLA
  • SupportDedicated

FAQ

Common questions

CogniSpend stores token counts, cost metadata, timestamps, and the attribution tags you set (workflow, team, use case). It does not store prompt text or model responses. Your content stays in your infrastructure.

The CogniSpend proxy is an OpenAI-compatible API endpoint. Point your existing client's base URL at the proxy, pass your provider API key as normal, and the proxy forwards the request to the real provider while recording token usage and cost. Round-trip latency added is typically under 5ms.

No. The SDK wraps your existing client and records usage data asynchronously after each response. It does not sit in the request path. Tracking data is batched and sent in the background so your agents run at full speed.

Both the SDK and proxy support streaming. For streaming requests, token counts are accumulated as chunks arrive and the final cost is recorded when the stream closes — your streaming latency is unaffected.

Yes. Any provider that exposes an OpenAI-compatible API (Groq, Together AI, Fireworks, Azure OpenAI, etc.) works through the proxy without changes. For providers with non-standard APIs, you can record costs via the CogniSpend REST API directly.

Enterprise plans include a self-hosted deployment option for organizations with data residency requirements. Contact us to discuss your setup.

One event equals one API call to a language model — one completion, one chat response, one embedding. Streaming responses count as one event regardless of the number of chunks. Manual cost entries do not count toward event volume.

The CogniSpend SDK is open source (MIT licensed) and available on GitHub. The core platform is commercial software. We publish the SDK so you can audit exactly what data leaves your system.

Both are tracked as fixed costs via manual cost entries or CSV import. You set the monthly seat cost, the number of seats, and the team it belongs to. CogniSpend attributes this spend alongside your API costs so you get a true total AI cost picture.

You can export all your data as CSV at any time from the settings page. On downgrade, data older than the new plan's retention window is archived for 30 days before deletion. On cancellation, you have 30 days to export before data is permanently removed.

Know what your AI is costing. Know if it's worth it.

Set up in under ten minutes. No infrastructure changes required. Start with your existing OpenAI or Anthropic client.

Free plan available. No credit card required to start.