The Best LLM Monitoring Tools in 2026, Ranked for Production Cost and Reliability

Monitoring an LLM app in production is a different job from debugging one in dev. In dev you want deep traces. In production you want three things - to know what it costs, to know when it breaks, and to not have your monitoring bill or your monitoring tool become the thing that breaks. I have run all five of these against live traffic, and the ones that failed failed on exactly those production realities, not on features.

Four axes, weighted for a running system:

• Cost and token visibility. Can you see spend per call, per model, per user, in real time.
• Self-host reality. Do you keep the real features when you run it yourself, and under what license.
• Pricing at volume. Whether the meter stays sane, or your monitoring goes dark past a cap.
• Reliability posture. Whether the tool adds a failure point to your request path.

The short version

Tool	Best for	Self-host	Cost tracking	Starting price
Langfuse	The all-round open-source default	Free, MIT, near-complete	Per trace	Free / $29/mo
Opik	Cheapest managed cloud	Free, Apache-2.0, full	Per span	Free / $19/mo
Arize Phoenix	OTel-native monitoring + RAG eval	Free, ELv2 server	Per trace	Free (OSS)
Portkey	Cost and routing across providers	Gateway only, not analytics	Native, enriched	Free / $49/mo
Helicone	Nobody new - maintenance mode	Free, Apache-2.0, frozen	Per request	Free / $79/mo

1. Langfuse - the default for most teams

Langfuse wins because it is the rare open-source tool where self-hosting gets you the real product. Only three features are enterprise-gated in the self-host build - tracing, evals, prompt management, human annotation and RBAC are all free under MIT. It tracks cost and token usage per trace, runs as an OpenTelemetry backend, and is framework-agnostic.

The economics seal it for a production budget. At 1M events a month it runs about $101/mo managed, against LangSmith’s roughly $2,514/mo for comparable volume - the widely-cited ~25x gap - and self-hosting removes the per-trace cost entirely.

The honest catch is operational. The v3 self-host is Postgres plus ClickHouse, Redis and S3-compatible storage - four services - and the migration to that architecture is where self-hosters get stuck. gRPC OTLP is not supported yet, HTTP only. And it is a ClickHouse subsidiary after the January 2026 acquisition. If you can run the stack, nothing else gives you this much for free; if you cannot, the $29/mo Core tier sidesteps it.

2. Opik - the cheapest managed monitoring

If you want production monitoring without running four services, Opik is the move. It is Comet’s platform, and the OSS build is Apache-2.0 with the full feature set self-hosted - unlimited spans, members and retention, no gates. On the cloud, Pro is $19/mo for 100k spans, the cheapest paid tier of the major platforms, with $5 per additional 100k. It captures multi-step traces with cost tracking, adds online (production) evaluation and guardrails, and it is the fastest-growing project of its peers at roughly 20.8k GitHub stars.

The gotcha is per-seat pricing at scale. The $19 headline is the small-team configuration, and seat costs climb as the team grows. A few users report UI slowdown on very large projects, and default cloud retention is 60 days - longer costs extra. For a small-to-mid team that wants managed monitoring cheap, it is hard to beat.

3. Arize Phoenix - OTel-native monitoring, best RAG eval

Arize Phoenix is the pick if you are standardizing on OpenTelemetry. It is genuinely OTel-native, built on OTel and Arize’s own OpenInference conventions, so it monitors any framework, and it runs locally in under a minute. Its 50+ pre-built eval metrics include the best RAG evaluation in the category, so you monitor answer quality, not just latency and cost.

Two gotchas. The server repo is Elastic License 2.0 - source-available, not OSI open source, and it forbids offering Phoenix as a hosted service to third parties (only the client and eval subpackages are Apache-2.0). And there are reports of ingest lag before traces appear, which matters if you need real-time production visibility. For internal monitoring the license rarely bites, and Arize is well-funded - a $70M Series C in February 2025.

4. Portkey - monitor cost and routing across providers

Portkey is a different animal. It is an LLM gateway that routes to 1,600+ models with fallbacks, caching, budgets and 50+ guardrails, and it happens to include monitoring. If your production problem is “we call five providers, costs are exploding, and we need budgets and fallbacks in one place,” it is aimed squarely at you. It ingests OpenTelemetry and enriches traces with cost and token metrics per the GenAI conventions - the strongest cost story in this set.

Know the split before you deploy. The open-source gateway is Apache-2.0 and self-hosts free - but that gets you routing and a basic dashboard, not real monitoring. Logs, traces, analytics and retention live on the managed Production tier at $49/mo for 100k logs. And the meter caps logs, not requests - past the cap your traffic keeps flowing but your monitoring quietly goes dark. Overage is $9 per additional 100k requests. Plenty of teams run Portkey as the gateway and feed a dedicated backend over OTel.

5. Helicone - do not start here

Helicone was a clean, Apache-2.0, proxy-based tool with fast cost tracking. Then Mintlify acquired it on 3 March 2026 and put it in maintenance mode - security patches, bug fixes and new-model support only, no roadmap, and Mintlify is actively helping customers migrate off. For a new buyer that overrides everything.

There is also a production reliability concern independent of the acquisition. The proxy model puts Helicone in your request hot path - a single point of failure. If Helicone is down, your LLM calls fail even when the provider is healthy, and every proxied call adds latency (Helicone cites ~10ms). An async SDK logging mode avoids both, at the cost of the zero-code convenience. If you are already on it, plan your exit. If you are evaluating fresh, skip to Langfuse or Opik.

So which one?

• You want the best all-round open-source monitoring and can run the self-host stack - Langfuse, roughly 25x cheaper than LangSmith at scale.
• You want managed monitoring for the least money - Opik at $19/mo, Apache-2.0.
• You are standardizing on OpenTelemetry, or you need the best RAG eval - Arize Phoenix, for internal use.
• Your real problem is cost and routing across providers - Portkey, and expect to pay for the real analytics.
• You are on Helicone today - plan your exit. Do not start there fresh.

Every price and date here was read from each vendor’s own materials and links to our full tool reviews. This category ships breaking changes monthly, so we re-verify every 30 days. For production, the thing that bites is rarely a missing feature - it is a meter that goes dark at volume or a tool that adds a failure point. Pick for the running system, not the demo.