Crypto Wallet Development Cost in 2026: Custody, MPC and Security Audits

In this article
- In 2026 a crypto wallet costs about $40,000–120,000 as an embedded wallet inside an existing app, $60,000–150,000 as a basic non-custodial mobile wallet, $150,000–350,000 as a multi-chain wallet with swaps and dApp connections, and $200,000–500,000 as a custodial wallet on a custody partner. Your own custody infrastructure starts around $500,000.
- Who holds the keys decides the budget. A non-custodial wallet is mostly app and security engineering. A custodial wallet adds KYC, a ledger, a compliance back office and, for US users, money transmission rules (general information, not legal advice).
- MPC and smart accounts (ERC-4337, and EIP-7702 since Ethereum's May 2025 Pectra upgrade) fix the seed phrase problem but add vendor fees, sponsored gas and contract audits. Buy key management from a provider unless custody is your product.
- Budget $15,000–100,000+ for pre-launch security testing and plan 15–20% of the build cost per year for maintenance: chains, SDKs and mobile OS releases change constantly.
Jump to
- The short answer: what a crypto wallet costs
- The cost axis that matters: who holds the keys
- Five wallet types, priced
- Seed phrase, MPC or smart account: the cost trade-offs
- Where the hours go in a multi-chain wallet
- Calculate your wallet budget
- Security: what a wallet must budget for
- Custodial wallets and US rules
- Hidden costs nobody puts in the first estimate
- What we see in estimates and first calls
- Which wallet type fits your plan?
- How to reduce the cost without weakening the wallet
- Team rates: who builds it and what it costs
- Where Gilzor fits
The short answer: what a crypto wallet costs
These are 2026 development costs for a US company working with a vendor at Central or Eastern European or Latin American rates, including QA, DevOps and project management. A US onshore team lands at roughly twice these numbers. A custodial wallet with its own custody infrastructure (MPC nodes or HSMs, signing services, withdrawal policies) typically starts around $500,000 and can pass $1.5 million. Security testing, licensing and running costs are separate lines, and we price them below.
This page is about wallets: products that hold, send and sign. If you are building a trading venue, the crypto exchange development cost guide covers order books, liquidity and exchange licensing. If the wallet is one part of a wider dApp (staking, NFTs, tokens), the blockchain app development cost guide prices the smart contracts and the audits around them.
The cost axis that matters: who holds the keys
Every wallet decision follows from one question. If the user holds the key, you are building security-sensitive software. If you hold it, you are running something regulators may treat as a financial service. In between sit MPC and smart accounts, which share control and move part of the cost from engineering to vendor fees.
Five wallet types, priced
1. Embedded wallet inside an existing app: $40,000–120,000
Your fintech, game, loyalty or marketplace app gets a wallet that users never have to install separately. Login with email, a social account or a passkey creates the wallet in the background through an MPC or smart account provider SDK. You build the wallet screens, transaction signing prompts, gas sponsorship rules, a recovery flow and the backend glue. The low number assumes one EVM chain and the provider's standard flows. Costs rise with custom recovery, several chains or your own smart account modules that need an audit.
2. Basic non-custodial wallet: $60,000–150,000
A standalone iOS and Android app (or a browser extension) for one or two chain families: create or import a wallet, back up a seed phrase, send, receive, balances, transaction history, address book, price data and push notifications. It sounds small. The hours go into secure key storage (iOS Secure Enclave and Keychain, Android Keystore), biometric locking, fee estimation, handling stuck or replaced transactions, and an indexer or data API so history loads in under a second instead of scanning the chain.
3. Multi-chain non-custodial wallet: $150,000–350,000
Several chain families (EVM networks, Bitcoin, Solana and others), token and NFT display, in-app swaps through an aggregator, a fiat on-ramp, staking, and dApp connections through WalletConnect or an in-app browser. Each non-EVM chain family is its own integration: different address formats, signing, fee models and failure modes. A Bitcoin UTXO wallet and an Ethereum account wallet share a logo, not code. In our estimates, every extra chain family adds roughly 120–250 hours before testing.
4. Custodial wallet on a custody partner: $200,000–500,000
Users sign up with KYC, fund the wallet by card or bank transfer, and hold crypto that a regulated custodian or crypto-as-a-service provider actually holds. You build onboarding, a double-entry internal ledger, deposits and withdrawals with risk checks, statements, support tooling and an admin back office for compliance and operations. Most of the money goes to the ledger and the back office, not the wallet screens. The partner model is also how many US fintechs add crypto without first getting their own licenses; the fintech app development cost guide covers the money movement side.
5. Custodial wallet with your own custody infrastructure: $500,000–1.5 million
You run the keys: MPC nodes or HSM-backed signing, hot, warm and cold tiers, withdrawal policies with multi-person approvals, key generation ceremonies, address whitelisting, real-time balance reconciliation per chain and monitoring that pages someone at 3 a.m. This is the wallet stack inside exchanges and institutional custodians. Build it only when custody is the product. Even large crypto businesses often license MPC custody technology rather than write threshold signature code themselves.
Seed phrase, MPC or smart account: the cost trade-offs
Key management is the decision that most changes both the build and the monthly bill. Here is how the options compare from a budget point of view.
| Approach | Added build effort | Running cost | Audit need | Best for |
|---|---|---|---|---|
| Seed phrase on device | Low: 100–200 h for secure storage, backup and import flows | Near zero | App pen test | Crypto-native users who expect self-custody |
| MPC via provider SDK | Low to medium: 150–300 h | Provider fee per monthly active wallet, usually with a free tier, then tiered pricing | App pen test, provider due diligence | Mainstream users, multi-chain products |
| Smart accounts (ERC-4337 / EIP-7702) | Medium: 300–600 h with paymaster, bundler and passkeys | Bundler and paymaster services, sponsored gas per transaction | Contract audit if you write or modify modules | EVM apps that want gasless onboarding, batching, spending limits |
| Own MPC / TSS implementation | High: 1,200–2,500 h plus cryptography expertise | Your own node infrastructure and on-call | Specialist cryptography review, often $100k+ | Custody companies, where key management is the product |
| Custodian API | Low on keys, high on ledger and compliance | Custody fees, often basis points on assets plus minimums | Partner due diligence, app pen test | Custodial fintech wallets |
Two dates explain why smart accounts are now a mainstream budget item rather than an experiment. ERC-4337 brought account abstraction to Ethereum without a protocol change when its EntryPoint contract went live in March 2023. Then Ethereum's Pectra upgrade on May 7, 2025 activated EIP-7702, which lets an ordinary account delegate to smart contract code, so existing users get batching and gas sponsorship without moving to a new address. More than 11,000 EIP-7702 authorizations were recorded in the first week after Pectra, according to on-chain data reported by The Block. The catch: delegation is also a new phishing target, so the signing prompts for it need real design and testing time.
Where the hours go in a multi-chain wallet
Take a mid-sized non-custodial wallet: iOS and Android on a shared codebase, a browser extension, four chain families, swaps, an on-ramp and WalletConnect, around 3,800 hours. Clients expect the screens to dominate. They don't.
The pattern is familiar from every product that moves money: the happy path is the small part. The large part is a transaction that sits in the mempool for an hour, a node that returns a stale nonce, a token with a non-standard contract, a user who switches networks mid-signature, a phone that kills the app during a background sync. QA on a wallet means testing each of those states on each chain, on real devices.
Calculate your wallet budget
Pick the wallet type, platforms, chains and key management. The calculator estimates development hours with QA and project management, converts them at the regional rate you choose, and adds pre-launch security testing and a first year of running costs.
Crypto wallet: development and year-one cost
Hours are typical medians from our estimates, not a quote. Provider fees use rough assumptions ($0.08 per monthly active wallet for MPC, $0.12 for smart accounts incl. sponsored gas on a layer 2); KYC assumes 10% of active wallets are new each month at about $1.85 per check. Licensing and compliance staff are not included.
Try switching the wallet type from multi-chain non-custodial to a custodial partner model. Hours rise less than you might expect, but the running cost and the work outside the calculator (legal, partner due diligence, compliance staff) change the whole economics. The other big lever is key management: building your own MPC adds more hours than swaps, staking, NFTs and dApp connections combined.
Built by Gilzor
Results we’ve shipped




Talk to the people who build it. Tell us about your project and get a free estimate of scope, timeline and cost.
Security: what a wallet must budget for
Wallets are where attackers go for volume. According to Chainalysis' 2026 Crypto Crime Report, $3.4 billion was stolen from crypto services and users in 2025. Losses from individual wallets fell to about $713 million from roughly $1.5 billion a year earlier, but the number of personal wallet compromises rose to more than 158,000 incidents affecting at least 80,000 victims. Attackers are taking smaller amounts from far more people, mostly through phishing, malicious signatures and malware on the device.
The most expensive wallet failures of recent years were not broken cryptography. They were ordinary engineering mistakes around the keys:
- Secrets in logs. In August 2022, thousands of Solana wallets were drained after the Slope mobile wallet was found to have sent seed phrases to a centralized logging service. One debug setting turned a wallet into a key database.
- Weak key generation. A flaw in the Profanity vanity address tool made generated keys recoverable; the market maker Wintermute lost about $160 million through one such address in September 2022.
- Supply chain. In December 2023, a compromised release of Ledger's Connect Kit library injected a wallet drainer into dApps that loaded it, and users lost hundreds of thousands of dollars within hours.
So the security budget for a wallet is a set of concrete line items:
- External pen test and code review of key generation, storage, signing, networking and update mechanisms: typically $15,000–60,000 per round, plus a fix and re-test round.
- Smart contract audit if you deploy smart account modules, paymasters or recovery contracts: often $20,000–80,000.
- Cryptography review for any custom MPC or threshold signature code, which can exceed $100,000 on its own.
- Transaction simulation and clear signing: showing users what a signature actually does (token approvals, EIP-7702 delegations) is now table stakes. Build it or pay a simulation API.
- Dependency and release hygiene: pinned and reviewed SDKs, signed builds, and a bug bounty after launch.
Our QA team treats signing and recovery flows as money paths: every state, every chain, real devices. Our internal benchmark is that only 5% of tasks sent to QA come back to developers, and on wallet work we add test cases rather than relax that bar. If you already have a wallet in production and want an outside view, a mobile app audit is a cheaper first step than a rebuild.
Custodial wallets and US rules
This is budgeting context, not legal advice. In its 2019 guidance on convertible virtual currency, FinCEN said that providers of unhosted wallet software, where users control their own keys, are generally not money transmitters. The guidance looks at who owns the value, where it is stored, whether the owner interacts with the network directly, and whether the provider has independent control over the funds. A wallet where your company can move user funds sits on the other side of that line.
That has engineering consequences. A custodial wallet usually needs FinCEN registration as a money services business with a Bank Secrecy Act AML program, state money transmitter licenses or a licensed partner's coverage, KYC at onboarding, sanctions screening, transaction monitoring, and Travel Rule data on transfers of $3,000 or more. Each of those becomes software: onboarding flows, case management, holds and reviews, immutable audit logs. The state-by-state licensing picture and its costs are covered in our crypto exchange development cost guide. Apple's App Store guidelines also matter for distribution: they allow apps to facilitate virtual currency storage only when the developer is enrolled as an organization, and both Apple and Google ask for licensing documents in many markets for exchange-like features.
Hidden costs nobody puts in the first estimate
| Cost | Typical size | When it hits |
|---|---|---|
| Maintenance and upgrades | 15–20% of the build cost per year | Chain upgrades and hard forks, SDK and wallet standard changes, yearly iOS and Android releases |
| RPC nodes and data APIs | From a few hundred dollars to several thousand a month | Grows with chains, active wallets and history lookups; spikes on volatile days |
| Key management provider | Fee per monthly active wallet after a free tier, or an annual contract | From launch, scaling with users; check exit and key export terms |
| Sponsored gas | Cents per transaction on layer 2s, far more on Ethereum mainnet | Every gasless transaction you promise; cap it per user or abuse will find it |
| Security testing | A pen test round per major release, audits for every new contract | Yearly at minimum |
| Support and recovery | Staff time for lost devices, failed recoveries and stuck transactions | From the first week; poor recovery UX shows up as support tickets |
| Custodial extras | Custody fees, KYC at about $1.35–1.85 per check at published rates, blockchain analytics, compliance staff | Monthly, growing with volume |
| App store fees | Apple's $99 yearly developer program, Google's $25 one-time fee; review delays on crypto features | Every release |
What we see in estimates and first calls
A few things come up often enough that we ask about them before putting a number on paper:
- The custody model is not decided. "Non-custodial, but we want password reset" is a custodial or MPC product. Deciding this first changes the architecture, the vendors and the legal track. A short business analysis phase pays for itself here.
- "Support all the chains." Each chain family is integration, nodes, QA and a new class of support tickets. Launch on the chains where your users already hold assets and add the rest when usage justifies it.
- Seed phrases for mainstream users. If the audience has never used crypto, a 12-word backup screen at signup costs you more in drop-off than an embedded MPC or passkey wallet costs in fees.
- History is an afterthought. Reading balances and transactions straight from nodes looks cheap in a demo and gets slow and expensive at scale. Budget an indexer or a data API from the start.
- Real-time state breaks on mobile. Websocket price and balance updates have to follow the app lifecycle. When we worked on the Android app of the KickEX crypto exchange, getting the socket connection right meant reworking the architecture, and fixing that kind of problem in an app that is already live is never cheap. Our tech troubleshooting team sees the same issue in inherited wallet apps.
Which wallet type fits your plan?
Five questions about your users, the custody model and the budget. The result points to the wallet type and the cost band that goes with it.
Find your crypto wallet type
How to reduce the cost without weakening the wallet
- Decide the custody model firstUser-held, shared through MPC or smart accounts, or custodial. It sets the architecture, the vendors and whether a compliance track exists at all.
- Buy key management, build the experienceMPC and smart account providers have spent years on key security. Your money goes further on onboarding, signing clarity and recovery UX.
- Launch on fewer chainsTwo well-supported chain families beat eight with flaky history and stuck transactions. Add chains when users ask for them.
- Use a shared codebase for iOS and AndroidCross-platform cuts app cost noticeably for wallet apps, as long as key storage uses each platform's secure hardware through native modules. Our mobile app development cost guide has the trade-offs.
- Integrate swaps and on-ramps, don't build themAggregators and ramp providers handle liquidity, payments and their own compliance, usually for a fee share.
- Start with an MVP scopeSend, receive, history and one killer feature. Our MVP development cost guide covers how to cut scope without cutting quality.
Skipping the external pen test, or shipping custom signing or recovery code without review, to hit a launch date. A slow roadmap is recoverable. A wallet that leaks keys usually isn't. If the budget is tight, cut chains, NFTs or staking, not security work.
Team rates: who builds it and what it costs
The same 3,000-hour multi-chain wallet lands at very different prices depending on who builds it. Using 2026 vendor rates for a senior-weighted team:
| Region | Typical vendor rate | 3,000-hour build | Overlap with US hours |
|---|---|---|---|
| US onshore | $120–160/h | $360,000–480,000 | Full |
| Latin America (nearshore) | $50–75/h | $150,000–225,000 | 6–9 hours |
| Central & Eastern Europe (offshore) | $50–75/h | $150,000–225,000 | 2–4 hours with the East Coast on a shifted schedule |
| South & SE Asia (offshore) | $30–50/h | $90,000–150,000 | 0–2 hours |
For wallet work, ask every vendor how they store keys on each platform, how they test signing on each chain, and who reviews their security before an external auditor does. Rate differences matter less than those answers. Our nearshore software development rates guide breaks rates down by country.
FAQ
How much does it cost to develop a crypto wallet in 2026?
What is cheaper to build: a custodial or a non-custodial wallet?
How long does it take to build a crypto wallet?
Should we use MPC or smart accounts for our wallet?
How much does a crypto wallet security audit cost?
What does it cost to run a crypto wallet after launch?
Where Gilzor fits
We build and stabilize web and mobile products for startups and SMBs from Poland and Cyprus, including fintech and crypto apps, through our mobile app development and web development services. For US clients we're an offshore team with two to four overlap hours with the East Coast. We don't give legal advice, so on custodial products we work alongside your counsel and custody partner. Every wallet estimate we send names the custody model, the chains, the key management vendor and the security testing it includes, so you can compare it line by line with any other quote.
No sales pitch
Get a straight answer for your project
Tell us what you’re building. We’ll reply with options, a rough cost and timeline. If we’re not the right fit, we’ll say so.

CTO of Gilzor. Responsible for architecture and the engineering standards our teams work by.
LinkedIn →Gilzor · Mobile Development partner
Need a team for your mobile app?
Services
Mobile DevelopmentNative and cross-platform iOS and Android apps, from MVP to scale.→By company type
Selected projects






The team behind them





