ModelTrace ObservatoryModelTrace
Pricing

Start free, scale with confidence

Transparent pricing that grows with your AI infrastructure. No hidden fees, no surprise overages.

Starter
For teams getting started with AI observability. Everything you need to instrument your first models.
Free
  • Up to 10,000 traces / month
  • 7-day trace retention
  • Drift detection for 2 models
  • Basic alerting (email)
  • Community support
  • 1 team member
Most Popular
Pro
For growing teams that need deeper visibility and faster incident response across their AI stack.
$149/month
  • Up to 500,000 traces / month
  • 30-day trace retention
  • Drift detection for unlimited models
  • Intelligent alerting (Slack, PagerDuty, webhooks)
  • Evaluation metrics & scoring
  • Trace replay & comparison
  • Priority email support
  • Up to 10 team members
Enterprise
For organizations with advanced compliance, scale, and integration requirements.
Custom
  • Unlimited traces
  • 90-day+ trace retention
  • Custom evaluation pipelines
  • SSO / SAML authentication
  • Dedicated support engineer
  • SLA guarantees (99.95% uptime)
  • On-prem / VPC deployment options
  • Unlimited team members

Full feature comparison

FeatureStarterProEnterprise
Monthly traces10,000500,000Unlimited
Trace retention7 days30 days90+ days
Model monitoring2 modelsUnlimitedUnlimited
Drift detection
Safety scoring
Evaluation metrics
Trace replay
Custom eval pipelines
Slack / PagerDuty alerts
SSO / SAML
VPC deployment
SLA guarantee
Team members110Unlimited
SupportCommunityPriority emailDedicated engineer

Frequently asked questions

What counts as a trace?

A trace is a single end-to-end request through your AI pipeline. This includes the prompt, any retrieval or chain-of-thought steps, and the final model output. Multi-step agents count as one trace per top-level invocation.

Can I try Pro features before committing?

Yes. Every new account gets a 14-day free trial of Pro features. No credit card required. At the end of your trial, you can choose to upgrade or continue with the Starter plan.

How does billing work for overages?

We never surprise you with overage charges. If you approach your trace limit, we notify you and let you choose to upgrade or stay on your current plan. Traces beyond your limit are sampled, not dropped.

Do you offer discounts for annual billing?

Yes. Annual plans save you 20% compared to monthly billing. Contact our sales team for annual pricing on Pro and Enterprise tiers.

What integrations are supported?

ModelTrace integrates with OpenAI, Anthropic, Cohere, open-source models via LiteLLM, LangChain, LlamaIndex, and any custom pipeline via our REST API or Python/Node SDKs.

Is my data secure?

Absolutely. All data is encrypted in transit and at rest. Enterprise customers can deploy in their own VPC. We are SOC 2 Type II certified and GDPR compliant.

Ready to see your AI clearly?

Start with our free tier and upgrade when your team is ready for deeper observability.