Datadog LLM Observability review
LLM tracing and evaluation with Datadog context, plus a free plan and cloud-only deployment.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
Datadog LLM Observability, now presented as Agent Observability, helps engineering teams monitor and troubleshoot LLM-powered applications and agents in development and production. It records model calls and workflows at span level, including inputs, outputs, latency, errors, token usage, and metadata. That makes it most relevant to teams that need both quality review and operational visibility, particularly those already working in Datadog.
Its feature set covers more than tracing: teams can run managed or custom evaluations, build datasets from production traces, compare prompts, models, and agent configurations in experiments, and annotate outputs for review. Dashboards and monitors track LLM metrics, while sensitive-data scanning and redaction cover application inputs and outputs. The notable fit is the connection between agent activity and backend services, infrastructure, and user sessions. Integrations include OpenAI, Anthropic, Gemini, Vertex AI, AWS Bedrock, LangChain, CrewAI, Pydantic AI, LiteLLM, and Strands Agents; instrumentation is available through SDKs and an HTTP API.
The Free plan includes up to 40K LLM spans per month, 15-day trace retention, unlimited context and evaluations, and full feature access. Pro is $160 per month with annual billing and includes 100K spans per month; additional spans cost $3.50 per 10K spans per month on annual billing. Both plans list 15-day retention and full feature access. Since deployment is cloud-only, teams requiring another deployment model should look elsewhere. Datadog is a strong fit for teams seeking LLM monitoring alongside their existing observability context; buyers who prioritize deployment flexibility should consider alternatives.
Datadog LLM Observability pros and cons
- Where it wins
- Links agent traces with backend services, infrastructure, and user sessions
- Combines evaluations, datasets, experiments, and annotations
- Free plan includes up to 40K LLM spans per month and full feature access
- Where it doesn't
- Cloud-only deployment
- Free and Pro plans both retain traces for 15 days
- Pro costs $160/month with annual billing; extra spans are metered
Datadog LLM Observability fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update