LangSmith pricing

About

LangSmith is the commercial observability and evaluation platform from LangChain — the company behind the eponymous open-source LLM framework (one of the most-starred AI projects on GitHub) and the LangGraph agent-orchestration library. LangSmith gives teams tracing, debugging, evaluation, and prompt/dataset management for LLM and agent applications, and works whether or not you build on the LangChain framework.

LangChain raised a $25M Series A in February 2024 (led by Sequoia, ~$200M valuation), the same month LangSmith reached general availability. In October 2025 it closed a $125M Series B led by IVP at a $1.25B valuation — minting it a unicorn — with strategic participation from Databricks, Datadog, Cisco, ServiceNow, and Workday. Revenue is widely understood to come primarily from LangSmith seats and trace usage plus LangGraph Platform deployments.

For current rates, see LangSmith’s pricing page.

Pricing summary : How LangSmith’s pricing model works

LangSmith is a hybrid model: a per-seat subscription that gates access and rate limits, plus usage-based per-trace billing for the work itself. There are three tiers — Developer ($0, 1 seat, 5k base traces/mo), Plus ($39/seat/mo, unlimited seats, 10k base traces/mo), and Enterprise (custom, with self-hosted and hybrid deployment). Beyond each plan’s included traces you pay as you go.

The distinctive part is how traces are metered. A “trace” is one run tree for a single request through your app. Traces are billed in two retention classes:

• Base traces — 14-day retention — $2.50 per 1,000.
• Extended traces — 400-day retention — $5.00 per 1,000.

You don’t pick a retention class up front for every trace. Every trace incurs the base charge; a trace then auto-upgrades to extended (a separate +$2.50/1k line on your invoice) when you add feedback, route it into an annotation queue, or an automation rule matches it. So the things that make observability useful — human review, eval scoring, automations — are exactly what doubles the per-trace cost.

What makes this different: most observability tools meter on a single event/span volume. LangSmith stacks two multipliers (seats × trace volume) and then bifurcates the usage axis by retention, so the same trace can be billed twice once it earns feedback.

Pricing by product

Per-trace overage (both Developer and Plus):

Sales motions across products: self-serve / PLG for Developer and Plus (credit card, instant signup), sales-led for Enterprise (self-hosted, hybrid, custom retention, procurement). LangChain also sells LangGraph Platform (agent deployment/hosting) and meters Deployment Runs ($0.005 each on Plus) on the same pricing page — related but separate SKUs; this entry focuses on LangSmith observability + evals. For a primer on metering choices like these, see choosing the right usage metric.

Hidden costs : What LangSmith users actually pay

The sticker price ($39/seat) understates real bills because two axes compound. Worked example for a five-person team running a moderately busy multi-step agent on Plus, ~200k traces/month, retained for audit (extended):

Two things bite teams that aren’t in the headline number:

1. Seat creep. Teams budget for 5 developer seats; then QA wants to inspect traces, PMs want to read failure cases, and data science wants to score outputs. A jump from 5 to 12 seats is +$273/mo nobody planned for.
2. The extended-retention upgrade. Every trace you add feedback to, queue for annotation, or hit with an automation rule silently moves from $2.50 to effectively $5.00 per 1k. High-eval workflows pay the upgrade on a large share of traffic.

At scale the trace axis dominates: published analyses peg ~500k base traces/mo at ≈$1,225/mo in overage, 1M at ≈$2,475/mo, and 5M (not unusual for high-traffic agents) at ≈$12,475/mo — before seats or extended-retention upgrades.

Want to estimate your own LangSmith bill? Use the LangSmith pricing calculator to model seats and trace volume.

Pricing evolution : LangSmith pricing history and changes

Cadence

Tracked range: 2024 GA → present. The $39/seat Plus price has been remarkably stable since launch; the notable evolution is on the usage axis — the two-retention-class trace model and the auto-upgrade mechanic. A Wayback snapshot of the pricing page existed (2026-05) but failed to render during capture, so historical-snapshot metadata is intentionally left empty rather than guessed.

Notable changes

• 2024-02 — LangSmith reaches GA with the $0 / $39-seat / custom three-tier structure.
• 2025-10 — Series B; LangSmith folded into a broader observe-evaluate-deploy platform narrative; pricing mechanics unchanged.
• 2026-06 — Verified base $2.50/1k (14-day) and extended $5.00/1k (400-day) trace pricing.

What’s unique : LangSmith’s distinctive pricing mechanics

1. Two retention classes on the usage axis. Rather than one trace price, LangSmith splits billing into base (14-day, $2.50/1k) and extended (400-day, $5.00/1k). Retention — usually a backend concern — is turned into the primary usage price lever.

2. Auto-upgrade on engagement. A trace jumps from base to extended the moment you add feedback, annotate it, or an automation matches it. The pricing literally tracks which traces you care about, monetizing the high-value subset.

3. Two compounding multipliers. Seats gate access; traces meter usage. Both grow with team adoption, so cost scales on adoption and traffic simultaneously — generous to a solo developer (free), expensive for a busy production team.

Strengths & weaknesses

Billing UX : LangSmith billing controls and transparency

• Usage limits — Organizations can set monthly caps on total traces and on extended-retention traces specifically. Limiting is approximate (cached, ~1–2 min to apply), so it’s a soft guardrail, not a hard stop.
• Rate limits — POST/PATCH to /runs* and /feedbacks* are capped at 5,000/min; plan-level hourly event limits run from 50,000 (Developer, no card) up to 500,000 (Plus/Startup).
• Usage visibility — A two-line invoice separates “Traces (Base Charge)” from “Traces (Extended Data Retention Upgrades),” which makes the upgrade cost auditable after the fact — though it’s only visible after the spend.
• Payment options — Credit card self-serve for Developer/Plus; invoicing and procurement on Enterprise. Data region (US/EU/APAC on GCP) is chosen at signup and can’t be changed later.

Strategic wins : Why LangSmith’s pricing decisions worked

1. A free tier that rides an enormous OSS funnel

LangChain’s framework brought millions of developers; the free Developer tier converts that distribution into LangSmith signups with near-zero friction. The OSS framework is the top of the funnel, LangSmith is the monetization layer. See how AI companies structure pricing.

2. Metering on retention, not just volume

By charging more for the traces you keep around (and care about), LangSmith aligns price with value — the traces feeding evals and audits are the ones worth paying for. Related: choosing the right usage metric.

3. Stable seat price, flexible usage

Holding Plus at $39/seat since GA gives buyers an anchor they trust, while the per-trace axis quietly captures upside as usage grows — a clean separation of “predictable” and “elastic” revenue. See outcome-based pricing trends.

Areas to improve : Gaps in LangSmith’s pricing approach

1. Seat tax on the wrong people

Charging $39 for every QA tester, PM, or analyst who just wants to read traces pushes teams to ration access — the opposite of what an observability tool should encourage. A cheaper read-only seat would defuse the most common complaint. See bill shock and cost unpredictability.

2. Hard-to-forecast extended-retention upgrades

Because traces upgrade automatically on feedback/annotation/automation, teams can’t easily predict what fraction of traffic will hit the $5/1k rate. Pre-spend estimation tooling (not just post-hoc invoice lines) would help.

3. Price gap vs. usage-only rivals at scale

Langfuse and similar tools undercut LangSmith dramatically for high-volume teams with no per-seat multiplier. As trace volumes climb, the seat tax plus dual-retention pricing becomes a real defection risk for cost-sensitive buyers.

Key takeaways

1. Hybrid by design. LangSmith stacks per-seat subscription on usage-based per-trace billing — predictable seats, elastic traces.
2. Retention is the price lever. Base (14-day, $2.50/1k) vs extended (400-day, $5.00/1k), with traces auto-upgrading the moment they earn feedback or annotation.
3. Two multipliers compound. Real bills scale with seats and traffic, so a tool that’s free for a solo dev gets expensive fast for a production team.
4. The OSS framework is the funnel. A free Developer tier monetizes LangChain’s massive open-source distribution.
5. Scale invites defection. At high trace volumes, per-usage rivals are several-fold cheaper — the seat tax is the soft underbelly.

UBP implications

1. Metering on retention is a clever value-alignment move — charge more for the data customers choose to keep — but it demands strong forecasting tooling or it becomes bill-shock fuel.
2. Per-seat + per-usage hybrids must watch the seat axis. When the people who need read access aren’t the people generating value, seat pricing actively discourages the adoption you want.
3. An OSS top-of-funnel plus a usage-metered commercial layer is a durable AI-infra pattern — but only if the paid metric (here, traces) stays defensible against cheaper single-axis competitors.

Sources

• LangSmith / LangChain pricing page (accessed 2026-06-09)
• LangSmith administration & billing concepts (accessed 2026-06-09)
• Fortune — LangChain $125M Series B at $1.25B valuation (accessed 2026-06-09)
• LangChain blog — Series B announcement (accessed 2026-06-09)
• The Hacker News — AgentSmith LangSmith security disclosure (accessed 2026-06-09)

Bottom line

LangSmith monetizes LangChain’s enormous open-source distribution with a clean hybrid model: a free Developer tier, a stable $39/seat Plus plan, and usage-based per-trace billing split into two retention classes ($2.50/1k base for 14 days, $5.00/1k extended for 400 days). The design aligns price with value — you pay more for the traces you keep and review — but two compounding multipliers (seats × traces) plus an auto-upgrade mechanic make real bills hard to forecast and leave the door open to cheaper usage-only rivals at scale.

Want to compare LangSmith against other LLM observability and MLOps companies? Browse the pricing blueprint.