LangSmith reprices usage into LCU/LSU units; base traces double to 0.005 LSU
LangSmith replaced its per-1k trace dollar rates with normalized LCU/LSU units. Base traces now cost 0.005 LSU each — double the old $2.50/1k — and extended retention drops to 180 days.
Base traces $2.50 / 1,000 (14-day retention); extended traces $5.00 / 1,000 (400-day retention), upgrade +$2.50 / 1,000. Deployment billed per deployment run ($0.005) plus uptime ($0.0036/min production, $0.0007/min development). Fleet billed per run (50 runs/mo on Developer; 500 included on Plus, then $0.05/run). Sandboxes billed in dollars per vCPU-hr / GiB-hr.
All usage metered in two normalized units — LCU (LangChain Compute Unit, $1.50) and LSU (LangChain Storage Unit, $1.00). Base trace 0.005 LSU; extended-retention upgrade 0.0025 LSU with retention cut to 180 days. Deployment charged on resources consumed (runtime compute 0.045 LCU/vCPU-hr, runtime memory 0.006 LCU/GiB-hr, database compute 0.177 LSU/vCPU-hr, database memory 0.025 LSU/GiB-hr) with 1 free Serverless (Small) deployment on Plus. Fleet metered in LCU (5 LCU/mo free on Developer, 25 LCU/mo on Plus) with model costs now included. Sandboxes re-denominated into LCU/LSU at the same effective dollar rates. Seats unchanged: $0 Developer, $39/seat Plus.
Proof of change
LangSmith's pricing pages, as we captured them on two dates.
Showing the whole page as captured — scroll either panel, or open it at full size.
Show 1 other page we compared
Showing the whole page as captured — scroll either panel, or open it at full size.
- 3 pages had no counterpart in the earlier capture and are not shown.
Both images are our own captures, taken on the dates shown. The pixels are unmodified — nothing is retouched, and where a region is highlighted the marker is drawn over the image, not into it. Open either capture to see it at full resolution.
LangChain rebuilt the LangSmith pricing page around two normalized meters — LangChain Compute Units for work and compute, LangChain Storage Units for traces and storage — and stopped publishing dollar-per-1,000-trace rates entirely. The per-trace rate is now disclosed only in a hover tooltip on the feature table: 0.005 LSU per base trace, 0.0025 LSU per extended-trace upgrade.
The re-denomination is not cost-neutral everywhere. Sandbox rates convert back to exactly their previous dollar values (0.0384 LCU x $1.50 = $0.0576 per vCPU-hr, the old published rate), which confirms the conversion factors — and by the same arithmetic base traces moved from $2.50 to $5.00 per 1,000, a 100% increase. The extended-retention upgrade held at its prior effective rate, so an all-in extended trace rose from $5.00 to $7.50 per 1,000. Extended retention itself was shortened from 400 days to 180 on the pricing page, while the billing docs still say 400 — a live inconsistency.
Deployment and Fleet also moved off per-run pricing onto resource metering. The LangSmith billing docs note that the new model “replaces the previous per-run and uptime pricing”, with existing customers grandfathered until October 1, 2026, and scale-to-zero available only on the new pricing. The page also gained a LangSmith Usage Calculator that estimates a monthly bill across Observability & Evals, Engine, Deployments, Sandboxes and Fleet.
LangSmith replaced its published dollar-per-1k trace rates and Deployment/Fleet per-run pricing with two normalized meters: LangChain Compute Units (LCU, $1.50) and LangChain Storage Units (LSU, $1.00). Base traces are now 0.005 LSU each (double the prior $2.50/1k), the extended-retention upgrade is 0.0025 LSU, and extended retention shortened from 400 days to 180. Deployment and Sandboxes moved to resource metering; Fleet moved from per-run to LCU and now includes model costs. Seats unchanged at $0 / $39. Existing deployment customers grandfathered until 2026-10-01.