LLM observability tools compared: 10 platforms
Prices and features checked 2026-10-17 on each vendor's own pages. “not verified” means we could not confirm it.
Short answer: To keep traces on your own servers for free, self-host Langfuse, Arize Phoenix, Comet Opik or Helicone. For a hosted service, the published entry plans start at $19/month (Comet Opik Pro Cloud) and free allowances include 5k base traces (LangSmith Developer), 25k spans (Arize AX Free, Comet Opik Free Cloud) and 50k units (Langfuse Hobby). Check retention: free tiers keep data for 3 to 60 days.
| Platform | OpenTelemetry | Self-hosting | Meter | Retention | Free tier | Entry paid price |
|---|---|---|---|---|---|---|
| Arize AX and Phoenix | OpenTelemetry compliant on all AX plans [source] | Phoenix self-hosts; AX self-hosted only on Enterprise [source] | Trace spans and ingestion volume (GB) [source] | 15 days (AX Free), 30 days (AX Pro), custom on Enterprise [source] | AX Free: 25k spans/month, 1 GB ingestion, 15-day retention, unlimited users [source, checked 2026-10-17] | AX Pro $50/month: 50k spans, 10 GB, 30-day retention (overage rates not stated on the pricing page) [source, checked 2026-10-17] |
| Braintrust | Accepts OpenTelemetry traces and logs [source] | Self-hosted data plane; on-prem or hosted deployment on Enterprise [source] | Processed data (GB) plus scores, plus model credits at token rates [source] | 14 days (Starter), 30 days (Pro), up to 180 days at $0.50/GB/month [source] | Starter $0: 1 GB processed data, 10k scores, 14-day retention, unlimited users; then $4/GB and $2.50 per 1k scores [source, checked 2026-10-17] | Pro $249/month: 5 GB and 50k scores included, then $3/GB and $1.50 per 1k scores; 30-day retention [source, checked 2026-10-17] |
| Comet Opik | Native OpenTelemetry support [source] | Open-source version self-hosts with unlimited spans [source] | Spans per month [source] | 60 days on Free and Pro Cloud; unlimited when self-hosted; extension to 400 days at $29 per 100k spans [source] | Free Cloud: 25k spans/month, up to 10 team members, 60-day retention; open source self-hosted is free [source, checked 2026-10-17] | Pro Cloud $19/month: 100k spans, up to 50 members, then $5 per 100k spans [source, checked 2026-10-17] |
| DeepEval and Confident AI | Confident AI docs include an Integrations & OTEL section [source] | DeepEval runs locally; Confident AI on-prem on Enterprise [source] | Flat monthly fee per organization plus trace spans in GB-months [source] | Retention billed in GB-months; adjustable [source] | Free: 2 seats, 1 project, 5 test runs/week, 1 GB-month of trace spans [source, checked 2026-10-17] | Starter $200/month per organization: unlimited seats, 5 projects, 5 GB-months, then $1 per GB-month [source, checked 2026-10-17] |
| Galileo (Splunk Agent Observability) | Log with OpenTelemetry [source] | Observability Cloud and self-hosted deployments [source] | Spans (per 1.2 million spans), billed annually [source] | not verified | not verified | From $100/month per 1.2 million spans, billed annually; includes 500 million Luna tokens and 5 Luna fine-tuning runs per year [source, checked 2026-10-17] |
| Helicone | Async logging through OpenLLMetry [source] | Docker or Kubernetes self-hosting [source] | Requests and storage (GB) beyond the free allowance [source] | 7 days (Hobby), 1 month (Pro), 3 months (Team) [source] | Hobby: 10,000 requests/month, 1 GB storage, 1 seat, 7-day retention [source, checked 2026-10-17] | Pro $79/month plus usage-based pricing (usage rates are shown only via a calculator, not verified) [source, checked 2026-10-17] |
| Langfuse | OpenTelemetry ingestion on all plans [source] | Free to self-host (Docker Compose or Kubernetes) [source] | Billable units: traces, observations (spans, events, generations) and scores [source] | Data access: 30 days (Hobby), 90 days (Core), 3 years (Pro and Enterprise) [source] | Hobby: 50k units/month, 30 days data access, 2 users [source, checked 2026-10-17] | Core $29/month: 100k units included, then $8 per 100k units (graduated down to $6 per 100k above 50M) [source, checked 2026-10-17] |
| LangSmith | OpenTelemetry tracing via OTLP endpoint [source] | Self-hosted or hybrid on Enterprise only [source] | Seats plus base traces (14-day retention) and extended traces (180-day) [source] | Base traces 14 days; extended traces 180 days (extra fee) [source] | Developer $0: 1 seat, up to 5k base traces/month, then pay as you go [source, checked 2026-10-17] | Plus $39/seat/month: up to 10k base traces/month, then pay as you go (per-trace rate not stated on the pricing page) [source, checked 2026-10-17] |
| Maxim AI | Accepts OpenTelemetry (OTLP) traces [source] | In-VPC deployment on Enterprise [source] | Seats plus logs per month [source] | 3 days (Developer), 7 days (Professional), 30 days (Business) [source] | Developer: free for up to 3 seats, 1 workspace, 10k logs/month, 3-day retention, no overages [source, checked 2026-10-17] | Professional $29/seat/month: 100k logs/month, 7-day retention, $1 per 10k extra logs [source, checked 2026-10-17] |
| W&B Weave | Import OpenTelemetry traces [source] | Self-managed Weave instance [source] | Weave data ingestion (GB/MB) [source] | not verified | Free $0: 1 GB/month Weave data ingestion [source, checked 2026-10-17] | Pro from $60/month: 1.5 GB/month ingestion, then $0.10 per MB; for teams under 50 employees [source, checked 2026-10-17] |
What to check before you pick one
- Instrumentation: if you instrument with OpenTelemetry, you can switch or fan out to several backends later. Every platform above documents an OpenTelemetry path (Helicone through OpenLLMetry).
- Where data lives: self-host one of the open-source options, or check which plan adds self-hosting or data residency.
- Retention versus cost: longer history usually means a higher plan or a per-GB storage fee (Braintrust $0.50/GB/month above 30 days; Opik $29 per 100k spans to extend to 400 days).
- Evals on traces: if you want to score production traces, check LLM-as-judge and annotation support on the evaluation tools comparison.
- Volume: run your numbers through the pricing calculator.
FAQ
What is LLM observability?
Recording what an LLM app or agent did on each request: the prompt, the model calls, tool calls and retrieval steps (spans) grouped into traces, with tokens, latency and cost. Most platforms then let you score those traces with evals and turn them into test datasets.
Which LLM observability tools support OpenTelemetry?
Verified on a vendor page: Arize AX and Phoenix, Braintrust, Comet Opik, DeepEval and Confident AI, Galileo (Splunk Agent Observability), Helicone, Langfuse, LangSmith, Maxim AI, W&B Weave. Helicone documents OpenTelemetry-based logging through OpenLLMetry. With OpenTelemetry you can send the same spans to more than one backend.
Which can I self-host?
Free to self-host from public code: Langfuse (MIT core), Arize Phoenix (Elastic License 2.0), Comet Opik (Apache-2.0) and Helicone (Apache-2.0). Self-hosting is on Enterprise plans for Braintrust, DeepEval and Confident AI, Galileo (Splunk Agent Observability), LangSmith, Maxim AI; W&B documents a self-managed Weave instance.
How long do they keep traces?
On the checked date: Arize AX and Phoenix: 15 days (AX Free), 30 days (AX Pro), custom on Enterprise. Braintrust: 14 days (Starter), 30 days (Pro), up to 180 days at $0.50/GB/month. Comet Opik: 60 days on Free and Pro Cloud; unlimited when self-hosted; extension to 400 days at $29 per 100k spans. DeepEval and Confident AI: Retention billed in GB-months; adjustable. Helicone: 7 days (Hobby), 1 month (Pro), 3 months (Team). Langfuse: Data access: 30 days (Hobby), 90 days (Core), 3 years (Pro and Enterprise). LangSmith: Base traces 14 days; extended traces 180 days (extra fee). Maxim AI: 3 days (Developer), 7 days (Professional), 30 days (Business).
How is LLM observability priced?
By the unit each vendor counts: units (Langfuse), spans (Arize, Opik, Splunk), GB processed or ingested (Braintrust, Confident AI, Weave), traces plus seats (LangSmith), logs plus seats (Maxim) or requests (Helicone). Use the pricing calculator to compare on your own volume.
Is this a ranking?
No. Platforms are listed alphabetically with the same fields, and we publish no star ratings. Sponsored placements, if any, are labelled and never change order or facts.
Comparisons and alternatives
- Langfuse vs LangSmith
- Arize Phoenix vs Langfuse
- Comet Opik vs Langfuse
- Braintrust vs LangSmith
- Langfuse alternatives · LangSmith alternatives · Arize alternatives · Braintrust alternatives
Data: tools.json · tools.csv (CC BY 4.0). Vendors: sponsorship options.