Runtime vs LangChain

LangChain and Runtime at a glance

LangChain makes the most widely used open-source tools for building agents, and LangSmith, the platform to run them. The open-source side has three layers: LangChain for getting an agent running quickly with any model provider, LangGraph for low-level control of agent workflows, and Deep Agents, an opinionated harness for long-running agents with planning, memory, and subagents. LangSmith adds Observability, Evaluation, Deployment, Sandboxes for running agent-generated code, an LLM Gateway, and Fleet, a no-code builder (formerly Agent Builder) that lets teams across a company create agents and use them in Slack, Teams, and Gmail. LangChain reports more than 350 million monthly open-source downloads and more than 7,000 LangSmith customers. It is built for developers and AI engineering teams, with Fleet extending to the rest of the company.

Runtime is the AI agent harness for payment and fintech teams: an operating system for building and running many agents across the org. Anyone in payment ops, risk, compliance, finance, underwriting, onboarding, or support builds agents from their SOPs. Agents work on their own isolated computers, reach your ledger, processor, and bank portals through APIs, databases, MCP servers, and a browser, and stop for a person before anything moves money. It is built for operations leaders who want agents working real queues this quarter.

Good products, different jobs. LangChain gives builders the toolkit. Runtime is the finished harness for the teams who run the money.

How they differ

A toolkit for builders vs a harness for operators

LangChain is at its best in an engineer's hands. LangGraph gives precise control over state and branching, Deep Agents gives a strong default loop, and LangSmith shows exactly what an agent did and scores it. Fleet brings agent building to non-engineers, but it is a general-purpose builder. Connecting it to your ledger, processor, and bank portals, and deciding how it behaves around money, is your work.

Runtime is built around operations work from the start. The first agent is easy on any stack. Running agents on payment data needs isolated computers, scoped credentials, RBAC, approvals, an audit trail, evals, model routing, fallbacks, and multi-cloud deployment. An engineering team can spend two quarters on that plumbing before the first agent touches real data. Runtime gives it to you on day one, and works with agents you already have.

Agent computers and where they run

LangChain offers LangSmith as SaaS, in your cloud (BYOC), or self-hosted, with self-hosted and hybrid on Enterprise plans. Code execution runs in LangSmith Sandboxes.

Runtime gives every agent run its own isolated computer in your AWS, GCP, or Azure account, or fully self-hosted via Helm, and lets you bring your own sandbox provider. Agents use a real browser for portals with no API, a common situation with sponsor banks and processors.

Models and harnesses

LangChain is model-agnostic, and Fleet supports OpenAI, Anthropic, and Gemini models. The harness is LangGraph or Deep Agents.

Runtime routes across harnesses (Claude Code, Codex, OpenCode) as well as models, with fallbacks when a provider goes down. You can serve open-weight models in your own cloud for PCI and PII work, and card numbers and SSNs are stripped from prompts and logs.

Guardrails designed for money movement

LangChain supports human-in-the-loop interrupts in LangGraph, and Fleet can require approval for sensitive actions through an agent inbox, with admin and access controls. These are general mechanisms. Which payment actions need a human, and what evidence the approver sees, is yours to define and build.

Runtime starts agents read-only and requires approval before anything moves money: releasing a payout, applying a reserve, booking a ledger entry, replying to a sponsor bank. Every run is recorded end to end, from the trigger through every query, tool call, approval, cost, and result, so the record your sponsor bank or examiner asks for stays with you, ready to export.

How you get to production

LangChain is self-serve for developers, with enterprise support SLAs on custom plans.

Runtime pairs you with a forward-deployed AI engineer from a team that built payment and fintech infrastructure at Hulu's payments team at Disney, Finix, Modern Treasury, and BlackRock's AI quant group. The FDE maps your processes, builds the first agents with your team, sets up guardrails, and trains admins. When a process is solved, agents can turn it into a deterministic script in your repos, so engineers own it if it becomes mission-critical.

Where LangChain is stronger

• Open source and community. The largest agent framework ecosystem, with integrations for almost every model and tool.
• Fine-grained control. LangGraph lets engineers design exactly how state, branching, and retries work.
• Observability and evals. LangSmith tracing and evaluation are widely used, and work with agents built on other frameworks.
• Breadth. Horizontal by design: coding agents, support agents, research agents, internal tools.
• Low entry cost. The frameworks are free, and LangSmith has a free tier and a $39 per seat plan.

Pricing

LangChain's frameworks are open source and free. LangSmith has a free Developer plan (one seat, up to 5k base traces per month), Plus at $39 per seat per month (up to 10k base traces, one free small deployment), and custom Enterprise pricing with self-hosted and hybrid deployment, SSO, and RBAC. Usage beyond the included amounts is pay as you go, with services such as Fleet and Sandboxes metered in LangChain Standard Units at $1 each. Model usage is paid to your model provider. The engineering time to build payment workflows, approval rules, and audit export is not on the price list.

Runtime has public tiers: Free ($0, one session), Teams from $99 per seat per month, and Enterprise with custom pricing and self-hosting. The value to weigh it against is the work of the analysts you were about to hire.

Which should you choose

Choose LangChain if

• You have AI engineers who want to own the agent code and its control flow
• You are building agents into your product
• Tracing and evaluation across many frameworks is your priority
• You want a general no-code builder for knowledge workers across the company

Choose Runtime if

• Payment ops, risk, compliance, or finance teams need agents on real queues soon
• Approvals before money moves and an examiner-ready audit trail are requirements
• You run on more than one cloud, or want your own sandbox provider and models
• You want a forward-deployed engineer who knows payments, not a framework

You can also run both: engineers keep building product agents with LangGraph, and operations runs on Runtime. Rain replaced $250k in vendor spend with Runtime.

Frequently asked questions

What is the difference between LangChain, LangGraph, and LangSmith?

LangChain and LangGraph are open-source frameworks: LangChain for getting agents running quickly with any model, LangGraph for low-level control of agent workflows. Deep Agents is an open-source harness for long-running agents. LangSmith is the paid platform for observability, evaluation, deployment, sandboxes, and Fleet, its no-code agent builder.

What is LangSmith Fleet?

Fleet, formerly Agent Builder, is LangSmith's no-code agent builder. Teams create agents from chat or templates, use them in Slack, Teams, and Gmail, and can require approval for sensitive actions through an agent inbox, with admin and access controls.

Is LangChain good for fintech and payments?

LangChain is horizontal and widely used, including by financial companies, and LangSmith reports SOC 2 Type II, HIPAA, and GDPR compliance. Payment-specific workflows, approval rules for money movement, and examiner-ready audit exports are what your team builds on top.

Can Runtime run agents built with LangGraph?

Runtime works with agents you already have and runs them alongside agents your ops teams build from SOPs, with the same approvals, masking, and audit trail.

How much does LangSmith cost?

LangSmith has a free Developer plan with one seat and up to 5k base traces per month, a Plus plan at $39 per seat per month with up to 10k base traces, and custom Enterprise pricing with self-hosted and hybrid options. Usage beyond the included amounts is pay as you go.