AI & Data

AI Agent Observability: What to Monitor, What It Costs

10 min read

Your AI agent will not fail with an alert. It will fail quietly: a retrieval index goes stale, a model version changes under you, one prompt edit breaks a downstream step, and cost per request creeps up while quality creeps down. LLM and agent observability is how you catch that before a customer does. At a minimum it covers five things: full traces of every multi-step run, token and latency and cost per request, quality evals on the outputs, prompt and model version tracking, and guardrail hits. Below is what to instrument, what the tools cost, and an honest build-vs-buy call.

We’re gmware, a custom software development firm in Austin, TX with engineering centers in Bangalore and Mohali, India, and we build and monitor AI features inside operational software for mid-market companies. This is the guide we wish more teams read the week after they shipped an agent, not the month they realized it had been quietly wrong for a while.

What LLM and agent observability actually covers

Traditional monitoring answers “is the service up and fast.” Agent observability answers a harder question: “did the thing do the right work, at what cost, and why did it get weird on Tuesday.” Those are not the same job. A 200ms response can be confidently wrong, and your APM dashboard will show it as a green success.

Five signals do most of the work. Traces come first: a single agent request can fan out into a dozen LLM calls, tool invocations, and retrieval steps, and you need the whole tree, not just the final answer. Then the money and speed layer, token count, latency, and cost per request, because that is what turns into a bill and an SLA. Then quality, which is the part APM can’t touch: did the output answer the question, cite real context, stay on policy. Then version tracking, so when quality drops you can tell whether someone changed the prompt, the model, or the retrieval data. And guardrail hits, the count of blocked or flagged outputs, which is your early-warning system for prompt injection and off-the-rails behavior.

What to instrumentWhat it catchesWhy plain APM misses it
Full run tracesWhich step in a multi-step agent broke, and whyAPM sees one HTTP call, not the 12 LLM/tool calls inside it
Token + cost per requestThe bill quietly doubling, one expensive promptNo concept of tokens or per-request model spend
Latency per stepThe slow tool call hiding inside a “fast” responseAggregates to a single span; the internal breakdown is gone
Output quality (evals)Hallucination, off-policy answers, empty retrievalsNo idea whether the text is right, only whether it returned
Prompt + model versionRegression traced to a specific changeNot tracked; a prompt edit looks like nothing changed
Guardrail / safety hitsPrompt injection, jailbreaks, blocked outputsNo hook into content moderation or policy checks

Why Datadog-style APM is not enough on its own

Your existing APM is genuinely good at what it does. It will tell you the endpoint threw a 500, the p95 latency spiked, the pod is out of memory. Keep it. The gap is that it treats an LLM call as an opaque HTTP request with a status code, and an agent’s real failures live below that line.

Here’s the concrete version. An agent answers a support question in 900ms with a 200 status. APM: perfect. What actually happened: the retrieval step returned nothing, the model made up a refund policy that doesn’t exist, and the guardrail that should have flagged it wasn’t wired to the moderation endpoint. None of that is visible as latency or error rate. It’s only visible if you traced the run, evaluated the output, and logged the guardrail decision. That’s the whole argument for a separate LLM layer. It doesn’t replace APM; it sees the part APM was never built to see.

The quality piece deserves its own line, because it’s the one people skip. Hallucination isn’t a pass/fail flag you can bolt on later. It’s a rate you have to measure against live traffic, because a model that looks fine on your test set can degrade on the messy real inputs users actually send. The common pattern now is to run an evaluator model (LLM-as-a-judge) on a sample of production traces plus every trace a guardrail flagged. Langfuse and LangSmith both document sampling roughly 1 to 10 percent of live traces for online evaluation to keep the evaluator bill sane. Sample the rest weekly by hand if you’re small. More on that below.

What the LLM observability tools actually cost

The market splits cleanly into open-source-you-host and SaaS-you-pay. Open source (Langfuse, Arize Phoenix) means the platform is free but you run the infrastructure. SaaS (LangSmith, Datadog LLM Observability, Helicone, and the hosted tiers of Langfuse) means you pay per event but skip the ops. Here’s where the money lands, pulled from each vendor’s own pricing page in July 2026.

ToolModelFree tierPaid entryNotes
LangfuseOpen source + hosted50k units/mo (Hobby)Core $29/mo, 100k unitsSelf-host free; Pro $199/mo adds long retention
Arize PhoenixOpen sourceFree, self-hostedN/A (host it yourself)OpenTelemetry tracing + evals; Elastic License 2.0
LangSmithSaaS5k traces/mo (Developer)Plus $39/seat/mo, 10k traces$2.50 per extra 1k base traces
Datadog LLM ObservabilitySaaS (APM add-on)None on this SKU$160/mo, first 100k LLM spans (annual)$3.50 per extra 10k spans; folds into existing Datadog
HeliconeSaaS (proxy)10k requests/moPro $79/moSits as a proxy in front of your LLM calls

Two things about that table. First, the sticker price is not the real cost. Overage is. Langfuse charges $8 per extra 100k units, Datadog $3.50 per extra 10k LLM spans, LangSmith $2.50 per extra 1k traces. A chatty agent that fires ten LLM calls per user request burns through “included” volume fast, so do the multiplication on your own traffic before you sign anything. Second, “span” and “trace” and “unit” and “request” are not the same unit across vendors, which makes head-to-head price comparison genuinely annoying. Read each pricing page’s definition, not just the number.

Open-source or SaaS: which fits your team

The real fork isn’t features, it’s who runs the infrastructure. Both open-source tools here are legitimately capable; Phoenix and Langfuse both do full tracing and evals. The question is whether you’d rather pay in dollars or in your own engineers’ time.

Self-hosting makes sense when your traces contain data you can’t send to a third party, or your volume is high enough that per-event SaaS pricing would hurt. SaaS makes sense when you want to be watching traces this afternoon and don’t want to babysit another service. If you’re already all-in on Datadog, its LLM Observability folding into the same dashboards has real pull, even at the higher entry price, because your on-call already lives there.

When a spreadsheet beats a paid observability platform

Here’s an opinion we’ll defend: if you’re running one low-volume agent, a paid observability platform is premature, and structured logging plus a spreadsheet of failures will serve you better for the first few months.

The reasoning is boring and correct. At low volume you don’t have a statistics problem, you have a “read the actual failures” problem, and a human reading 30 bad transcripts learns more than any dashboard will tell you at that scale. Log every run as structured JSON: input, the full step trace, output, token cost, latency, and a pass/fail you tag by hand. Pipe failures into a spreadsheet. Review it weekly. You’ll find the top three failure modes in an afternoon, and you’ll have spent zero dollars and half a day of setup instead of onboarding a platform you’re not ready to use.

SignalReach for a real platformA logs + spreadsheet setup is fine
VolumeThousands of runs a dayDozens to low hundreds a day
TeamMultiple people need shared dashboardsOne or two people who read the logs
Failure rateYou can’t eyeball every failure anymoreYou can still read every bad run by hand
EvalsYou need automated LLM-as-judge scoringManual pass/fail tagging still keeps up
ComplianceAuditors want retained traces and access logsNo formal retention requirement yet

Switch when any row on the left becomes true. The tell is usually volume plus a second person who needs to see the data. Once you can’t read every failure by hand, or a teammate is asking “where do I look,” the manual setup has done its job and it’s time to buy or self-host. Don’t switch before then just because a vendor has a nice demo. The same discipline that separates the AI pilots that survive applies here: instrument the thing you can actually act on, not the thing that looks impressive in a screenshot.

How gmware keeps shipped AI features from drifting

We instrument the agent the same week we build it, because retrofitting traces onto a system already in production is how a two-day job becomes a two-week one. That means the six signals up top, tracing, cost, latency, quality evals, versioning, and guardrail hits, wired in from the first commit, not bolted on after something breaks. Our AI agents and LLM integration practice runs delivery from Austin with engineering in Bangalore and Mohali, which keeps senior eyes on your production traces during US hours without US-only burn rates.

And we’ll tell you when you don’t need a platform yet. If you’ve got one agent doing a few hundred runs a day, we’ll set up structured logging and a failure review instead of selling you an observability subscription you’ll outgrow the meaning of before you outgrow the tier. When you do scale past that, our broader AI and machine learning work covers the eval suites and drift monitoring that keep quality honest as traffic grows. If you’re still upstream of all this and just costing out the agent itself, our guides on what it costs to add AI to existing software and trimming your LLM bill are the place to start.

Tell us what your agent does and how many runs a day it handles, and we’ll give you a straight answer on whether to log, self-host, or buy, within 48 hours.

  • ai observability
  • llm monitoring
  • ai agents
FAQ

Common questions, answered

What is LLM and AI agent observability?
It's monitoring built for the way agents actually fail. Five signals do most of the work: full traces of every multi-step run, token and latency and cost per request, quality evals on the outputs, prompt and model version tracking, and guardrail hits. Traditional APM tells you the service is up and fast; observability tells you whether the agent did the right work and why it drifted.
Why isn't Datadog-style APM enough for AI agents?
APM treats an LLM call as an opaque HTTP request with a status code. An agent can answer in 900ms with a 200 status while the retrieval step returned nothing and the model invented a policy. None of that shows up as latency or error rate. You only catch it if you traced the run, evaluated the output, and logged the guardrail decision. Observability sees below the line APM stops at.
How much do LLM observability tools cost?
Open source is free to self-host: Langfuse and Arize Phoenix both cost $0 in license. SaaS entry pricing in July 2026: Langfuse Core $29/mo, LangSmith Plus $39/seat/mo, Helicone Pro $79/mo, and Datadog LLM Observability $160/mo for the first 100k spans. The number that actually grows is per-event overage past your included volume, so model your real request count first.
Should I use open-source or SaaS observability?
The fork is who runs the infrastructure, not features. Self-host (Langfuse, Phoenix) when your traces hold data you can't send to a third party or your volume makes per-event SaaS pricing hurt. Choose SaaS when you want traces this afternoon and don't want to run another service. If you already live in Datadog, its LLM layer folding into existing dashboards has real pull despite the higher entry price.
When is a spreadsheet better than a real observability platform?
When you're running one low-volume agent. At dozens to low hundreds of runs a day you don't have a statistics problem, you have a read-the-failures problem, and a human reading 30 bad transcripts learns more than a dashboard. Log each run as structured JSON, pipe failures to a spreadsheet, review weekly. Switch to a platform once you can't read every failure by hand or a teammate needs shared access.

See it on your own data.

Book a 30-minute discovery call and we'll walk through your use case.