Blocks.ai publishes this comparison and sells in this market. Every claim about another product links to a public source.

DIY vs LangSmith

Q: How do the costs compare?

LangSmith’s Developer plan has one free seat. Plus is $39 per seat a month; usage is metered in LangChain Standard Units, $1 each. For Enterprise, LangChain says to contact sales. DIY’s protocol specifications are openly licensed; build and running costs are not publicly documented.

Build it yourself (DIY)

Building agent connections and controls in-house from open protocols, existing infrastructure, and open-source tools

LangSmith

Platform for observing, evaluating, and deploying agents

DIY is not a product: you connect and govern agents yourself, on open protocols such as MCP and A2A, under your own policies.⁠^{Source 1, Source 2, Source 3} LangChain describes LangSmith as a platform for observing, evaluating, and deploying agents; its Deployment service runs agents on Agent Servers with A2A and MCP endpoints.⁠^{Source 4, Source 5, Source 6, Source 7}

By Blocks.ai teamLast checked 2 October 2026

Where each one sits

Six layers of running AI agents at a company, and what each product’s own public sources say it covers.

These aren’t the same kind of product

Where they overlap: Both cover running and tracing agents: a DIY build can use frameworks such as LangGraph and tracing such as OpenTelemetry, and LangSmith deploys agents and accepts OpenTelemetry traces.⁠^{Source 5, Source 8, Source 9, Source 10}
Where they differ: LangChain says Fleet builds agents without code for non-technical teams.⁠^{Source 4, Source 11} DIY is assembled from parts you choose, such as Istio for service identity and policy, and LangGraph for agents.⁠^{Source 8, Source 12}
Running both: LangSmith documents an observability-only setup, in which agents run wherever you already run apps and send traces to LangSmith.⁠^{Source 13}

Public sources · checked 2 October 2026

Offered
You build it
Not publicly documented

DIY

Build agents: You build itOpen-source frameworks like ADK⁠^{Source 14}
Host and run agents: You build itSelf-hosted, like LangGraph⁠^{Source 8}
Identity and access: You build itYour own identity provider⁠^{Source 15}
Registry and governance: You build itYour own agent registry⁠^{Source 16}
Traffic between agents, tools, and models: You build itYour gateway and service mesh⁠^{Source 17}
Agents across organizations: You build itA2A with API management⁠^{Source 9}

LangSmith

Build agents: OfferedFleet agent builder⁠^{Source 11}
Host and run agents: OfferedLangSmith Deployment runtime⁠^{Source 5}
Identity and access: OfferedFleet per-agent permissions⁠^{Source 18}
Registry and governance: OfferedCentralized registry in Deployment⁠^{Source 19}
Traffic between agents, tools, and models: OfferedFleet forwards MCP tool calls⁠^{Source 20}
Agents across organizations: Not publicly documented

At a glance

DIYLangSmith

What it isAn in-house build on open protocols such as A2A and MCP, which the A2A specification calls complementary.⁠^{Source 1, Source 2}LangChain describes a platform for observing, evaluating, and deploying agents.⁠^{Source 4} Its no-code part, Fleet, creates agents.⁠^{Source 11}

Where agents runWherever you run them: Pinterest optimized for MCP servers hosted in its internal cloud.⁠^{Source 21}Managed by LangChain on AWS and GCP (Plus plan or above), or self-hosted or hybrid on Enterprise.⁠^{Source 5, Source 22} Agents elsewhere can send traces.⁠^{Source 13}

Access controlA2A calls authorization logic implementation-specific.⁠^{Source 1} MCP makes authorization optional.⁠^{Source 23}Hosted deployments check a LangSmith API key by default, or your own auth handler.⁠^{Source 24} Role-based access control is an Enterprise feature; Cloud Enterprise adds SAML SSO.⁠^{Source 25, Source 26}

Audit and tracingA2A docs advise auditing significant events and tracing, for example with OpenTelemetry.⁠^{Source 9}A tracing project for every deployment.⁠^{Source 27} On Enterprise plans, audit logs of administrative and configuration actions, kept up to 400 days.⁠^{Source 28}

PricingA2A and MCP are openly licensed specifications, A2A under Apache 2.0.⁠^{Source 1, Source 2} Build and running costs are not publicly documented.Developer plan with one free seat.⁠^{Source 22} Plus is $39 per seat a month.⁠^{Source 22} Enterprise: contact sales.⁠^{Source 22} Usage is metered in LangChain Standard Units (LSUs), $1 each.⁠^{Source 22}

What each one is

Build it yourself (DIY)

DIY means connecting and governing agents without buying an agent platform: protocols such as MCP and A2A wired together in-house, existing infrastructure stretched, an in-house platform, self-hosted open source, or no central approach; Uber says that before its gateway, teams built integrations independently, duplicating infrastructure.⁠^{Source 17}

LangSmith

LangSmith Observability takes traces from agents, including ones that run outside LangSmith.⁠^{Source 13} LangSmith Deployment runs agents on infrastructure LangChain manages (Plus plan or above) or, on Enterprise, self-hosted or hybrid.⁠^{Source 5, Source 22} LangSmith Fleet builds agents without code.⁠^{Source 11}

The differences that matter

Tracing and audit
DIY
You build it: A2A’s docs advise distributed tracing, for example with OpenTelemetry, and Pinterest’s MCP servers use shared library functions that log inputs and outputs.⁠^{Source 9, Source 21}
LangSmith
Each deployment gets a tracing project automatically, and Fleet traces capture tool calls, decisions, and outputs, which can be searched and exported.⁠^{Source 27, Source 29}
LangSmith keeps base traces 14 days and extended traces 180 days.⁠^{Source 22} For customers who signed up after 3 August 2026, bulk export is Enterprise-only.⁠^{Source 30}
Access control
DIY
Each A2A server authorizes requests under its own policies, and MCP makes authorization optional; Uber’s gateway applies policies from Uber’s internal Access Control System.⁠^{Source 1, Source 17, Source 23}
LangSmith
Hosted deployments check a LangSmith API key by default; authorization comes from @auth.on handlers you write.⁠^{Source 24} Self-hosted Agent Servers have no default authentication.⁠^{Source 24}
Fleet, LangSmith’s no-code part, has per-agent sharing permissions (Clone, Run, and Edit) that can be layered and revoked at any time.⁠^{Source 18, Source 29}
Agents at other companies
DIY
A2A is designed for agents built by different companies on separate servers, and each server authorizes requests under its own policies.⁠^{Source 1, Source 31}
LangSmith
Agent Server speaks A2A with other A2A-compatible agents.⁠^{Source 6} Fleet can share agents with users or your whole workspace, and resources can’t be shared across workspaces.⁠^{Source 29, Source 32}
Both can use A2A, which is designed for agents from different companies.⁠^{Source 6, Source 31} For LangSmith, an access model for agents owned by another organization is not publicly documented.

For security teams

What a security review asks, answered from each vendor’s public documentation.

DIYLangSmith

Network exposureMCP servers on Streamable HTTP expose an HTTP endpoint, and A2A agents on HTTP use HTTPS URLs in production.⁠^{Source 1, Source 33}Agent Server exposes A2A and MCP endpoints.⁠^{Source 6, Source 7} Enterprise: private access to LangSmith Cloud via AWS PrivateLink or Private Service Connect.⁠^{Source 34}

IdentityIn A2A, identity is established at the HTTP layer.⁠^{Source 9} Keycloak’s MCP authorization support is Experimental (in preview) from MCP 2025-06-18.⁠^{Source 15}Cloud Enterprise adds SAML SSO and SCIM.⁠^{Source 25} Fleet agents act with fixed credentials or the user’s own.⁠^{Source 35}

Access changes and revocationA2A servers authorize requests under their own policies, with implementation-specific logic.⁠^{Source 1}Admins can deactivate or delete any member’s personal access token, which stops working within a minute.⁠^{Source 32} Fleet sharing is revocable.⁠^{Source 29}

Audit trailA2A docs advise auditing task creation, critical state changes, and agent actions.⁠^{Source 9} MCP documents trace context propagation.⁠^{Source 36}Enterprise audit logs: administrative and configuration actions, kept up to 400 days.⁠^{Source 28} A scheduled hourly pull can feed a SIEM.⁠^{Source 28}

ComplianceA2A docs say to ensure compliance with regulations such as GDPR, CCPA, and HIPAA; MCP leaves data protections to implementers.⁠^{Source 3, Source 9}LangChain says its SOC 2 Type II report covers all of LangSmith.⁠^{Source 37} It also lists ISO 27001:2022, GDPR, and HIPAA.⁠^{Source 37}

Full comparison

18 criteria in five groups. Every cell links to its source, or says no public source answers it.

DIY and LangSmith compared on 18 criteria
What it is
	DIY	LangSmith
What it is and who it’s for	An in-house build on protocols such as A2A, an open standard for communication between agent systems, and MCP, which the A2A specification calls complementary.⁠^{Source 1}	LangChain calls it a framework-agnostic platform to observe, evaluate, and deploy agents.⁠^{Source 4} It says its no-code Fleet is for non-technical teams.⁠^{Source 4, Source 11}
Maturity	Varies by component: MCP’s latest specification revision is 2026-07-28.⁠^{Source 3} The official MCP Registry is in preview and may have breaking changes or data resets.⁠^{Source 38}	LangChain announced GA of LangGraph Platform (now Deployment) on 14 May 2025 and Agent Builder (now Fleet) on 13 January 2026.⁠^{Source 19, Source 39, Source 40, Source 41} Self-hosted Fleet is in beta.⁠^{Source 11}
Control
Agent registry and discovery	Build your own: A2A prescribes no standard API for curated registries.⁠^{Source 16} Pinterest’s internal registry is its source of truth for approved MCP servers.⁠^{Source 21}	LangChain says Deployment manages agents in a centralized registry with versioning and rollbacks.⁠^{Source 19, Source 42} Listing agents hosted elsewhere is not publicly documented.
Identity and access control	A2A leaves identity to the HTTP layer and calls authorization logic implementation-specific.⁠^{Source 1, Source 9} MCP authorization is optional.⁠^{Source 23}	API key by default on hosted deployments, or your own auth handlers.⁠^{Source 24} Fleet: per-agent Clone, Run, and Edit permissions.⁠^{Source 18} Enterprise: RBAC, and SAML SSO on Cloud.⁠^{Source 25, Source 26}
Ownership, policy, and revocation	MCP says implementers should build consent and authorization flows.⁠^{Source 3} At Pinterest, non-experimental MCP servers need an owning team, registry entry, and review.⁠^{Source 21}	Fleet agent cards show the owner, sharing can be revoked at any time, and pausing an agent’s channels pauses it.⁠^{Source 29, Source 41, Source 43} Enterprise: custom roles.⁠^{Source 26}
Audit log and observability	A2A docs advise auditing significant events and tracing, for example with OpenTelemetry.⁠^{Source 9} MCP documents trace context propagation.⁠^{Source 36}	A tracing project per deployment, and OpenTelemetry traces accepted.⁠^{Source 10, Source 27} Enterprise audit logs of admin actions, kept up to 400 days, in OCSF format.⁠^{Source 28}
Connection
How agents connect	MCP servers on Streamable HTTP expose an HTTP endpoint, and A2A agents on HTTP use HTTPS URLs in production.⁠^{Source 1, Source 33} Uber’s gateway calls via its service mesh sidecar.⁠^{Source 17}	Agent Server has an A2A endpoint.⁠^{Source 6} For deployment updates, the data plane polls the control plane, which never connects to it directly.⁠^{Source 27}
Agents across organizations	A2A is designed for agents from different companies on separate servers, each applying its own policies.⁠^{Source 1, Source 31} Agent Cards may be signed with JWS.⁠^{Source 1}	Agent Server talks to other A2A-compatible agents.⁠^{Source 6} Fleet can share with users or your whole workspace; resources can’t be shared across workspaces.⁠^{Source 29, Source 32}
Protocol support	MCP defines stdio and Streamable HTTP transports.⁠^{Source 44} A2A maps to JSON-RPC, gRPC, and HTTP/REST bindings.⁠^{Source 1}	Agent Server supports A2A, with documented gaps, over JSON-RPC only; MCP over Streamable HTTP.⁠^{Source 6, Source 7} Fleet discovers tools from remote MCP servers.⁠^{Source 20}
Frameworks, models, and clouds supported	A2A gives agents built on different frameworks, languages, or vendors a common language.⁠^{Source 1} Google’s ADK says it is model-agnostic and deployment-agnostic.⁠^{Source 14}	Deployment runs Google ADK, CrewAI, AutoGen, and other agents via a wrapper package or the LangGraph Functional API.⁠^{Source 45} Agents hosted elsewhere can send traces.⁠^{Source 13}
Operations
Deployment options and data residency	Wherever you run it: Pinterest optimized for MCP servers in its internal cloud.⁠^{Source 21} A2A’s docs leave protecting stored data to your own policies.⁠^{Source 9}	Cloud on GCP (US, EU, APAC) or AWS (US).⁠^{Source 34} BYOC on AWS, self-hosted, or hybrid, on Enterprise.⁠^{Source 13, Source 22, Source 46} Cloud region moves aren’t supported.⁠^{Source 47}
Compliance attestations	Sits with the implementer: A2A docs cite regulations such as GDPR, CCPA, and HIPAA, and MCP leaves access controls and data protection to implementers.⁠^{Source 3, Source 9}	Trust Center lists SOC 2 Type II, ISO 27001:2022, GDPR, and HIPAA.⁠^{Source 37} On the SaaS platform, data is encrypted with AES-256 at rest and TLS 1.2+ in transit.⁠^{Source 48}
Support and SLA	Depends on the component: MCP SDKs are tiered partly by maintenance commitments.⁠^{Source 49} The official MCP Registry, in preview, gives no uptime guarantees.⁠^{Source 38, Source 50}	Base, Standard, and Premium support plans.⁠^{Source 51} Reasonable commercial efforts toward 99.5% API uptime for SaaS, measured quarterly.⁠^{Source 51}
Time and effort to get running	A curated A2A registry is a service you deploy and maintain.⁠^{Source 16} Pinterest built a unified deployment pipeline after new MCP servers took too much setup.⁠^{Source 21}	Cloud deployment needs a Plus plan or above and an API key.⁠^{Source 52} Self-hosting the Deployment control plane needs Kubernetes and Enterprise.⁠^{Source 5}
Pricing model and public prices	A2A and MCP are openly licensed specifications, A2A under Apache 2.0.⁠^{Source 1, Source 2} Build and running costs are not publicly documented.	Developer: one free seat.⁠^{Source 22} Plus: $39 per seat a month, with one free small serverless deployment.⁠^{Source 22} Enterprise: contact sales.⁠^{Source 22} Usage metered in LSUs, $1 each.⁠^{Source 22}
Building
Agent building tools	Frameworks such as LangGraph and Google’s open-source Agent Development Kit build and deploy agents.⁠^{Source 8, Source 14} A2A has SDKs in six languages.⁠^{Source 53}	Fleet builds agents from a description or a template, without code.⁠^{Source 11} Code-first agents can be deployed with the LangGraph CLI and app templates.⁠^{Source 5}
Model access	Chosen by whoever builds the agents: Google’s ADK says it is optimized for Gemini and model-agnostic.⁠^{Source 14}	Fleet’s managed tiers (Fast, Pro, Max) may change models over time; custom models aren’t in that picker.⁠^{Source 54} LangChain says closed or open models work.⁠^{Source 4}
Integrations and ecosystem	The official MCP Registry, in preview, has a REST API for clients and aggregators to discover MCP servers.⁠^{Source 38}	Fleet has built-in tools for Gmail, Slack, Google Calendar, GitHub, and others, and its agents work in Slack, Teams, and Gmail.⁠^{Source 54, Source 55}

Which to choose

Choose DIY if

You want no agent platform contract, building on openly licensed specifications such as A2A and MCP.⁠^{Source 1, Source 2}
You want your service mesh or internal access-control system to check MCP calls, as at Uber and Pinterest.⁠^{Source 17, Source 21}
You want your own registry of approved MCP servers and your own review rules, as Pinterest built.⁠^{Source 21}
You want A2A over gRPC or HTTP/REST: the specification defines both, and LangSmith’s Agent Server implements JSON-RPC only.⁠^{Source 1, Source 6}

Choose LangSmith if

You want tracing for agents from LangChain or other frameworks, including agents you host yourself, through integrations or OpenTelemetry.⁠^{Source 10, Source 13, Source 56}
You want agents hosted for you on Plus or above, with A2A and MCP endpoints and versions you can roll back.⁠^{Source 5, Source 6, Source 7, Source 19, Source 42}
You want non-technical teams to build agents without code and share them with Clone, Run, or Edit permissions.⁠^{Source 4, Source 11, Source 18}
You want a vendor with one free Developer seat, published Plus prices, SOC 2 Type II, and Enterprise SSO and audit logs.⁠^{Source 22, Source 25, Source 28, Source 37}

Questions buyers ask

Is LangSmith an alternative to building it yourself?

For hosting and tracing agents, it can be: LangSmith Deployment runs agents in production, and each deployment gets a tracing project.⁠^{Source 5, Source 27} Listing agents hosted elsewhere, and access for agents owned by other organizations, are not publicly documented.

Does LangSmith support A2A and MCP?

Yes. Agent Server supports A2A, with documented gaps, over the JSON-RPC binding only, and can expose LangGraph agents as MCP tools over Streamable HTTP.⁠^{Source 6, Source 7} Fleet discovers tools from remote MCP servers.⁠^{Source 20}

Can LangSmith work with agents not built with LangChain?

Yes. LangSmith Deployment runs agents from frameworks such as Google ADK, CrewAI, and AutoGen through a wrapper package or the LangGraph Functional API.⁠^{Source 45} Agents hosted elsewhere can send traces to LangSmith, including over OpenTelemetry.⁠^{Source 10, Source 13}

How do the costs compare?

LangSmith’s Developer plan has one free seat.⁠^{Source 22} Plus is $39 per seat a month; usage is metered in LangChain Standard Units, $1 each.⁠^{Source 22} For Enterprise, LangChain says to contact sales.⁠^{Source 22} DIY’s protocol specifications are openly licensed; build and running costs are not publicly documented.⁠^{Source 1, Source 2}

How we compare

Read the full method

Every claim on this page links to a public source. Where none answers a question, the page says so.

We re-check every fact at least every 90 days. This page was last checked 2 October 2026.

Something wrong or out of date? Tell us and we’ll correct it.

Sources

60 public sources, each with the date we checked it. Every one opens in a new tab.

Source 1: Agent2Agent (A2A) Protocol Specification a2a-protocol.org · checked 2 Oct 2026 Back:a b c d e f g h i j k l m n o p q r s
Source 2: modelcontextprotocol/modelcontextprotocol LICENSE (GitHub) github.com · checked 7 Oct 2026 Back:a b c d e f
Source 3: Specification - Model Context Protocol 2026-07-28 modelcontextprotocol.io · checked 2 Oct 2026 Back:a b c d e
Source 4: LangSmith: The Agent Engineering Platform LangChain · checked 2 Oct 2026 Back:a b c d e f g
Source 5: LangSmith Deployment LangChain · checked 2 Oct 2026 Back:a b c d e f g h i
Source 6: A2A endpoint in Agent Server LangChain · checked 2 Oct 2026 Back:a b c d e f g h i j
Source 7: MCP endpoint in Agent Server LangChain · checked 2 Oct 2026 Back:a b c d e
Source 8: langchain-ai/langgraph README (GitHub) github.com · checked 2 Oct 2026 Back:a b c d
Source 9: Enterprise Features - A2A Protocol a2a-protocol.org · checked 2 Oct 2026 Back:a b c d e f g h i j k l
Source 10: Trace with OpenTelemetry LangChain · checked 2 Oct 2026 Back:a b c d e
Source 11: No-code agents with LangSmith Fleet LangChain · checked 2 Oct 2026 Back:a b c d e f g h
Source 12: Security - Istio concepts istio.io · checked 2 Oct 2026 Back to text
Source 13: Set up LangSmith LangChain · checked 2 Oct 2026 Back:a b c d e f g h
Source 14: google/adk-python README (GitHub) github.com · checked 2 Oct 2026 Back:a b c d
Source 15: Integrating with Model Context Protocol (MCP) - Keycloak keycloak.org · checked 2 Oct 2026 Back:a b
Source 16: Agent Discovery - A2A Protocol a2a-protocol.org · checked 2 Oct 2026 Back:a b c
Source 17: Designing MCP Gateway Uber's MCP Management Platform - Uber Blog uber.com · checked 2 Oct 2026 Back:a b c d e
Source 18: Agent platform comparison LangChain · checked 2 Oct 2026 Back:a b c d
Source 19: LangSmith Deployment LangChain · checked 2 Oct 2026 Back:a b c d
Source 20: Remote MCP servers LangChain · checked 2 Oct 2026 Back:a b c
Source 21: Building an MCP Ecosystem at Pinterest - Pinterest Engineering Blog medium.com · checked 2 Oct 2026 Back:a b c d e f g h
Source 22: LangSmith Plans and Pricing LangChain · checked 2 Oct 2026 Back:a b c d e f g h i j k l m n o p
Source 23: Authorization - Model Context Protocol specification 2026-07-28 modelcontextprotocol.io · checked 2 Oct 2026 Back:a b c
Source 24: Authentication & access control LangChain · checked 2 Oct 2026 Back:a b c d
Source 25: Authentication methods LangChain · checked 2 Oct 2026 Back:a b c d
Source 26: Role-based access control LangChain · checked 2 Oct 2026 Back:a b c
Source 27: LangSmith control plane LangChain · checked 2 Oct 2026 Back:a b c d e
Source 28: Audit logs LangChain · checked 2 Oct 2026 Back:a b c d e
Source 29: Access & oversight LangChain · checked 2 Oct 2026 Back:a b c d e f
Source 30: Bulk export trace data LangChain · checked 2 Oct 2026 Back to text
Source 31: a2aproject/A2A README (GitHub) github.com · checked 2 Oct 2026 Back:a b c
Source 32: Overview (administration) LangChain · checked 2 Oct 2026 Back:a b c
Source 33: Streamable HTTP - Model Context Protocol specification 2026-07-28 modelcontextprotocol.io · checked 2 Oct 2026 Back:a b
Source 34: Cloud (SaaS) LangChain · checked 2 Oct 2026 Back:a b
Source 35: Agent identity LangChain · checked 2 Oct 2026 Back to text
Source 36: Key Changes - Model Context Protocol specification 2026-07-28 modelcontextprotocol.io · checked 2 Oct 2026 Back:a b
Source 37: Trust Center - LangChain (update: 2026 SOC 2 Type II Report now Available) LangChain · checked 2 Oct 2026 Back:a b c d
Source 38: The MCP Registry - Model Context Protocol modelcontextprotocol.io · checked 2 Oct 2026 Back:a b c
Source 39: LangGraph Platform is now Generally Available: Deploy & manage long-running, stateful Agents LangChain · checked 2 Oct 2026 Back to text
Source 40: Now GA: LangSmith Agent Builder LangChain · checked 2 Oct 2026 Back to text
Source 41: LangSmith Fleet changelog LangChain · checked 2 Oct 2026 Back:a b
Source 42: Assistants LangChain · checked 2 Oct 2026 Back:a b
Source 43: Manage agent settings LangChain · checked 2 Oct 2026 Back to text
Source 44: Transports - Model Context Protocol specification 2026-07-28 modelcontextprotocol.io · checked 2 Oct 2026 Back to text
Source 45: Deploy other frameworks LangChain · checked 2 Oct 2026 Back:a b
Source 46: Bring Your Own Cloud (BYOC) LangChain · checked 2 Oct 2026 Back to text
Source 47: Regions FAQ LangChain · checked 2 Oct 2026 Back to text
Source 48: LangSmith shared responsibility model LangChain · checked 2 Oct 2026 Back to text
Source 49: SDK Tiers - Model Context Protocol modelcontextprotocol.io · checked 2 Oct 2026 Back to text
Source 50: MCP Registry Aggregators - Model Context Protocol modelcontextprotocol.io · checked 2 Oct 2026 Back to text
Source 51: Support Plans LangChain · checked 2 Oct 2026 Back:a b
Source 52: Deploy your app to cloud LangChain · checked 2 Oct 2026 Back to text
Source 53: A2A protocol roadmap a2a-protocol.org · checked 2 Oct 2026 Back to text
Source 54: Essentials LangChain · checked 2 Oct 2026 Back:a b
Source 55: LangSmith Fleet LangChain · checked 2 Oct 2026 Back to text
Source 56: LangSmith Observability LangChain · checked 2 Oct 2026 Back to text
Source 57: Your company's private network Blocks.ai · checked 6 Oct 2026 Back to text
Source 58: Network requirements Blocks.ai · checked 2 Oct 2026 Back to text
Source 59: Solutions: Agent sprawl Blocks.ai · checked 2 Oct 2026 Back to text
Source 60: What is Blocks? Blocks.ai · checked 2 Oct 2026 Back to text

Where each one sits

These aren’t the same kind of product

DIY

vsLangSmith

At a glance

What each one is

Build it yourself (DIY)

LangSmith

The differences that matter

Tracing and audit

Access control

Agents at other companies

For security teams

Full comparison

Which to choose

DIYChoose DIY if

LangSmithChoose LangSmith if

Questions buyers ask

How we compare

Sources

LangSmith

Choose DIY if

Choose LangSmith if