Agentic Index
The best AI agents for banking and financial services in 2026
Agentic Index tracks 135 AI agents for banking and financial services: 33 built for the industry and 102 broader platforms that also serve it. FLOWX.AI, Appian, and Moveo.AI document the broadest capabilities among the specialists. 19 of the 135 researched vendors publish their prices, and 112 sell by quote through a sales team.
AI agents for banking and financial services serve banks, credit unions, lenders, wealth managers and fintech companies. They answer customer questions, process loan documents, support advisors, review transactions for fraud and financial crime, and handle compliance checks.
Financial firms buy under tighter rules than most. Every vendor on this page names financial services as a customer. They differ most in the controls buyers in this industry ask for: audit trails, human approval, and deployment inside the firm's own environment.
Researched from public sources on the same 14 capabilities. No vendor pays for placement. How this list works
Where to start: 10 of 33 built for banking and financial services
Latest verified research: 2026-11-12
-
1FFLOWX.AIPriced by quote
Mortgage underwriting agents sit inside BPMN processes connected to Jack Henry, FIS, Finastra or Temenos cores, with an approval gate after each AI step. The AML and fraud stack gathers evidence into case files, builds event timelines and compiles SAR ready submissions for investigator review.
-
2AAppianPriced by quote
Customer Lifecycle Management runs onboarding, KYC and servicing as process models, and agents placed as steps inside them extract data from applicant documents with DocCenter for Financial Services and query customer records. Deterministic checks, validations and human approvals stay in the surrounding process, which keeps an audit trail of every run.
-
3MMoveo.AIPriced by quote
Collections agents reach out on WhatsApp, voice and chat, reference a customer's payment history and offer a repayment plan inside limits the bank sets. It also nudges new cardholders toward a first purchase, and it runs on premises with privately hosted models so conversations stay inside the bank.
-
4BBretton AIPriced by quote
AML, KYC, KYB and sanctions agents run a whole investigation for banks and fintechs, gathering evidence from the institution's records and more than 180 data sources and drafting the case to its policies. A compliance analyst approves each case, and agents work over systems the bank runs, such as NICE Actimize, nCino and FICO.
-
5AArtianPriced by quote
Break Remediation works breaks in trade and risk pipelines, Client Operations handles prospecting, onboarding and servicing for wealth firms, and Batch Operations catches operational incidents, each set to the firm's own rules. High impact fixes go to a person first, and every decision is kept in an audit log the firm can export.
-
6FForestPriced by quote
KYC and KYB onboarding, sanctions screening, AML investigations and dispute handling run as shared workflows where analysts and agents work the same case. The backend sits inside the bank or fintech's own infrastructure, high risk actions wait for human approval, and every step can be exported as a regulator ready bundle.
-
7GGliaPriced by quote
Routine calls and digital questions across more than a thousand banking journeys go to Glia Banker, and Agentic Workflows complete tasks such as freezing a card or moving money between accounts. Each institution picks which steps need a person, and Glia Outreach runs voice and SMS campaigns to grow deposits and cut delinquencies.
-
8SSardinePriced by quote
Agent Hub agents review sanctions and PEP alerts, monitor transactions, run business due diligence and file disputes, and each alert comes back with a recommended disposition and rationale for an analyst. The SAR agent drafts a narrative from the case summary only after the analyst decides to file.
-
9GGradient LabsPriced by quote
The Disputes Agent classifies a card dispute against scheme reason codes, gathers evidence and files the chargeback with Visa or Mastercard once a person approves, and the Lending Agent works borrowers from abandoned applications to collections. Each agent follows procedures written in plain English, and guardrails catch complaints and financial difficulty.
-
10OOpenseePriced by quote
Agensee works on Opensee's platform for sell side and buy side firms, covering market, counterparty credit and liquidity risk, collateral and margin, and IRRBB. Risk data teams have it check each day's data and flag anomalies in seconds, and a quant can describe a new calculator that it builds from the firm's methodology document.
Broader platforms that also cover banking and financial services
These 102 vendors sell a wider product and name work for banking and financial services as one of its uses. They suit a company that already runs on the platform or wants one agent layer across several teams.
- CCamunda
- The AI Agent sub process triages incoming loan files, calls credit and KYC checks as service tasks and hands borderline files to an underwriter in Tasklist. A gateway in front of the MCP client lets read operations through and sends write actions to a person for sign off.
- DDynamiq
- A lender's loan file review agent screens each upload with a PII guardrail, checks it against credit policy in the knowledge base and holds the update at an approval node for an underwriter to edit and approve. KYC and AML alert triage run the same way, with analyst approval and every step traced.
- PPega
- A bank's employees can call agents from chat or email while every action runs inside an approved, auditable workflow, so an agent cannot improvise its way through a regulated process. A document agent splits, sorts and triages complex customer files, and Agent Tracer records every step and response.
- UUiPath
- Agents extract onboarding data, run KYC and AML research across jurisdictions and route a complete due diligence package to the relationship manager, and they adjudicate sanctions and PEP alerts while analysts make the real risk calls. For card disputes they apply Reg E and network rules, with complex cases reviewed in Action Center.
- FFlowable
- Know Your Business onboarding runs as a case in which a document agent classifies incoming files, a knowledge agent checks them against policy and the orchestrator agent proposes the next step in the case chat. High stakes actions wait for an analyst, and every model exchange is logged down to the tokens used.
- GGrep AI
- Onboarding, credit and coverage work for banks and fintechs runs as agents built from a team's own review criteria, covering KYB, sanctions and adverse media screening, credit memos and client briefs. Swarms clear a backlog of cases and send flagged ones to analysts, and banks can deploy Grep inside their own perimeter.
Picks by what you need most
- Best for complex multi step work
-
FFLOWX.AI
Documents the most on orchestration, memory, and knowledge grounding among the specialists.
- Best for regulated and governed deployments
-
AAppian
Documents the most on security, auditability, and human oversight among the specialists.
- Best for developer extensibility
-
MMoveo.AI
Documents the most on APIs, MCP support, and tool calling among the specialists.
- Best for fastest time to value
-
BBretton AI
Documents the most on prebuilt packs and channel coverage among the specialists.
How they charge
- Published prices: 19 of 135
- Plans and prices are on the website, and a team can start without a sales call. Examples: GGrep AI,OOdin AI,SSimplAI andZZenML.
- Priced by quote: 112 of 135
- Sold through a sales team, with pricing set per customer. Examples: CCamunda,FFLOWX.AI,AAppian andDDynamiq.
- Pricing not published: 4 of 135
- No pricing model could be confirmed from public sources.
Where they run
The rest run as a cloud service operated by the vendor.
What to check before a shortlist
- Where it runs
- Many financial firms need private cloud or on premises deployment. Check the options before a long evaluation.
- What it may do alone
- Answering a balance question is different from moving money or approving a loan. Check which actions wait for a person.
- The trail it leaves
- Examiners ask for evidence. Look for a full record of what the agent saw, said and did.
- Which model it uses
- Model risk teams ask which model is behind an answer and how it is tested. Look at whether the firm can choose the model and review how it performs.
Head to head
All 135 AI agents for banking and financial services
Decagon
Lorikeet
Maisa
Sardine
Sierra
TraversalCommon questions
- What are the best AI agents for banking and financial services?
- FLOWX.AI, Appian, and Moveo.AI lead the 33 products built for banking and financial services tracked by Agentic Index, and Camunda, Dynamiq, and Pega lead the 102 broader platforms that also cover it. The list is ordered by how many of the same 14 capabilities each vendor documents in public sources, and no vendor pays for placement.
- How much do AI agents for banking and financial services cost?
- 19 of the 135 researched vendors publish their prices, so a team can see plans and start without a sales call. 112 sell by quote through a sales team. Each vendor profile on Agentic Index lists the pricing model and any published prices.
- Can AI agents for banking and financial services run in a private cloud or on premises?
- Most run as a cloud service. 33 of the 135 researched vendors document a private cloud option and 56 document an on premises or self hosted option.
- How many AI agents for banking and financial services are there?
- Agentic Index tracks 135 AI agents for banking and financial services. Every one is researched from public sources and listed on this page.
How this list works. A vendor is on this page when its product is built for banking and financial services, either as its main category in the index or as a job its own materials name. The order is the number of the 14 Agentic Index capabilities each vendor documents in public sources, with full support counting more than partial support and ties broken alphabetically. The picks, the pricing counts and the deployment counts are computed from the same records. No vendor pays for placement.