Start free, no credit card. Upgrade when your agents go to production.
Free
No credit card required
For trying Trodo and for side projects.
Most popular
$249/ month
Billed monthly
For teams running agents in production.
Need a free trial? Book a demo
Everything in Developer, plus:
Let's talk
For organisations with scale, security or deployment requirements.
Everything in Pro, plus:
| Feature | Developer | Pro | Enterprise |
|---|---|---|---|
| Usage limits | |||
| Units per month | 10,000 | 250,000 | Custom |
| Data retention | 7 days | 60 days | Custom |
| Evaluation scores per month | 1,000 | 50,000 | Custom |
| Issue investigations per month | Not included | 250 | Custom |
| AI Analyst questions per month | 5 | 100 | Custom |
| Max ingest rate | 60 units/min | 600 units/min | Custom |
| Max payload size | 512 KB | 1 MB | Custom |
| Observability | |||
| Distributed tracing, runs, spans, tool and LLM calls | Included | Included | Included |
| Token, cost and latency attribution | Included | Included | Included |
| Conversation and session timelines | Included | Included | Included |
| Semantic search across inputs and outputs | Included | Included | Included |
| Structured query builder and saved views | Included | Included | Included |
| Error and exception capture | Included | Included | Included |
| User and account attribution | Included | Included | Included |
| Run feedback capture | Included | Included | Included |
| Python, Node.js and browser SDKs | Included | Included | Included |
| OpenTelemetry ingest, run alongside Datadog or Jaeger | Included | Included | Included |
| Framework auto-instrumentation | Included | Included | Included |
| Data export | Included | Included | Included |
| Evaluations and experiments | |||
| Automated evaluators on live traffic | Included | Included | Included |
| LLM-as-judge and rule-based scoring | Included | Included | Included |
| Human-in-the-loop grading queue | Included | Included | Included |
| Evaluator backfill across historical runs | Not included | Included | Included |
| Datasets from production traces | Included | Included | Included |
| Experiments and regression tracking | Included | Included | Included |
| Score-based cohorts | Not included | Included | Included |
| Prompt studio | |||
| Playground and multi-panel model comparison | Included | Included | Included |
| Generation-quality scoring (7 dimensions) | Included | Included | Included |
| LLM-judge scoring and cross-model divergence | Included | Included | Included |
| Prompt versioning and full history | Included | Included | Included |
| Named labels and runtime resolution | Included | Included | Included |
| Tools and response formats versioned with the prompt | Included | Included | Included |
| Prompt deployment without a code release | Included | Included | Included |
| Custom model providers | Not included | Included | Included |
| Issues and capabilities | |||
| Failure pattern detection | Included | Included | Included |
| Signal surfacing on runs and spans | Included | Included | Included |
| Automated root-cause investigation | Not included | Included | Included |
| GitHub source-code investigation | Not included | Included | Included |
| Issue lifecycle and regression detection | Not included | Included | Included |
| Affected-user impact analysis | Not included | Included | Included |
| Capability discovery and clustering | Not included | Included | Included |
| Configurable detection thresholds | Not included | Included | Included |
| AI Analyst | |||
| Natural-language questions over your data | 5/month | 100/month | Custom |
| Multi-step reasoning with charts and tables | Included | Included | Included |
| Deep analysis for multi-part questions | Not included | Included | Included |
| Answers linked to source records | Included | Included | Included |
| Automation and alerting | |||
| Real-time alerting on issues and scores | Included | Included | Included |
| Email notifications | Included | Included | Included |
| Slack notifications | Not included | Included | Included |
| Visual workflow builder | Not included | Included | Included |
| GitHub, Jira, Linear and Salesforce actions | Not included | Included | Included |
| Scheduled and event-driven execution | Not included | Included | Included |
| MCP server access | Included | Included | Included |
| Workspace | |||
| Users | Up to 3 | Unlimited | Unlimited |
| Workspaces | 1 | Unlimited | Unlimited |
| Agents per workspace | Unlimited | Unlimited | Unlimited |
| Role-based access (owner, admin, member, billing) | Included | Included | Included |
| Shared allowance pooling across workspaces | Not included | Not included | Included |
| Security and deployment | |||
| Google SSO | Included | Included | Included |
| Custom SSO, your identity provider | Not included | Not included | Included |
| Role-based access control | Included | Included | Included |
| GDPR compliant | Included | Included | Included |
| Data Processing Agreement (DPA) | Not included | Included | Included |
| SOC 2 | In progress | In progress | In progress |
| Custom data retention policy | Not included | Not included | Included |
| Hosting | Multi-tenant SaaS | Multi-tenant SaaS | SaaS, single-tenant, self-hosted or BYOC container |
| Security review | Not included | Not included | Included |
| Support | |||
| Documentation and community | Included | Included | Included |
| Email support | Not included | Included | Included |
| Dedicated Slack channel | Not included | Not included | Included |
| Uptime and support SLA | Not included | Not included | Included |
| Onboarding and team training | Not included | Not included | Included |
One unit is one record stored. A run is 1 unit, each span within it is 1 unit, and each event is 1 unit, so an agent run with 27 spans costs 28 units. Only rows that actually store are counted, so a retried batch is never charged twice.
Ingestion for that metric is refused until your next monthly period, and the API returns a clear error naming the limit you reached and your current usage. Nothing is silently dropped, and your existing data is unaffected. Upgrading raises the limit immediately.
On your billing anniversary, not the first of the calendar month. If you signed up on the 18th, your period runs 18th to 18th.
No. The Developer plan is free forever, and the 14-day Pro trial does not require a card.
Yes, 14 days at full Pro limits. At the end of the trial you keep your data and drop back to Developer limits unless you upgrade.
Yes, at any time. Upgrades take effect immediately.
Install the Python, Node.js or browser SDK and initialise it with your project key. Most teams are sending traces within fifteen minutes. If you already run OpenTelemetry, point your existing exporter at Trodo, no new SDK required, and it runs alongside Datadog, Jaeger or Honeycomb rather than replacing them.
No. Trodo is built for the AI-specific layer, prompts, evaluations, agent reasoning and failure analysis, and is designed to sit alongside your infrastructure monitoring.
Data is logically separated per account, access is role-based, and we are GDPR compliant with a DPA available. SOC 2 certification is in progress. Enterprise customers can run Trodo self-hosted or in their own cloud (BYOC).
Yes, on Enterprise, self-hosted or as a BYOC container deployment in your own environment.
Annual billing on Pro is $2,499, two months free. Get in touch for startup pricing.