Embedded-wallet landscape 2026 · post-consolidation provider map

TL;DR

By mid-2026 the embedded-wallet category has compressed from a long tail of 20+ SDK vendors into a default-tier duopoly (Privy under Stripe, Coinbase CDP) and a secondary tier (Magic, thirdweb, Dynamic, Web3Auth, Particle Network, Crossmint) competing for niches the hyperscalers’ default routing does not cover. The driver is not product superiority — it is hyperscaler default-wallet placement: when AWS Bedrock AgentCore, Azure AI Foundry Agent Service, and Google Cloud Vertex AI Agent Builder each pick “default” SDKs, the chosen vendors capture the long tail of AI-agent provisioning at zero distribution cost. See Privy x AWS AgentCore default-wallet positioning for the canonical example and embedded wallet network effects · integrator moat for the underlying flywheel.

Wiki route

This entry sits under agent-economy index. Read it with Privy · Stripe subsidiary embedded self-custody wallet for the canonical Stripe-side example, Coinbase CDP · developer platform overview for the Coinbase-side example, embedded wallet integrator moat for the moat thesis, and embedded wallet fintech disintermediation for the fintech-side framing.

Default-tier vs secondary-tier split

The category split visible in public docs and hyperscaler announcements by mid-2026:

Tier	Provider	Owner	Hyperscaler default	Public footprint	Notes
Default	Privy	Stripe (acquired Jun 2025)	AWS Bedrock AgentCore (default), Azure AI Foundry (one of two defaults)	110M+ wallets, 1,000+ integrations	Routes through Stripe Connect + Bridge USDB rail (see [[fintech/embedded-wallet-fintech-disintermediation-stripe-trojan-horse
Default	Coinbase CDP	Coinbase (NASDAQ: COIN)	AWS Bedrock AgentCore (co-default with Privy), GCP Vertex AI Agent Builder (default)	~150K developers, ~8K dApps, ~5M MAU embedded wallets	Default-binds Base + USDC + Onramp
Secondary	Magic	Independent (Tiger / SV Angel)	Not a hyperscaler default	Long-tail Web3 SaaS clients	Magic Link / OAuth + DKLs MPC; pricing tiered per-MAU
Secondary	thirdweb	Independent	Not a hyperscaler default	Heavy on NFT / gaming / consumer dApps	InAppWallet + smart-account SDK; chain-agnostic; partial open source
Secondary	Dynamic	Independent (a16z crypto)	Not a hyperscaler default	DeFi / consumer-finance focus	Multi-chain login, embedded + external-wallet bridge
Secondary	Web3Auth	Independent (Torus Labs, BNB-Labs backed)	Not a hyperscaler default	Long-tail Web3 + enterprise	tKey + OAuth share reconstruction; open-source
Secondary	Particle Network	Independent	Not a hyperscaler default	EVM + BTC + multi-chain	Modular SDK + own chain-abstraction layer
Secondary	Crossmint	Independent	Not a hyperscaler default	NFT / commerce / RWA	Specializes in fiat-funded wallets + receipt rails

“Default” means the SDK appears in the hyperscaler’s first-class agent-orchestration documentation as the recommended path, with credentials / billing / IAM wiring pre-built. Developers can swap to any other SDK, but the default is what the bulk of net-new agents inherit.

The five drivers of consolidation

1. AWS Bedrock AgentCore default selection (2025-Q3 GA). AWS picked Privy + Coinbase CDP as the two default wallet providers for AgentCore Payments. Circle, Anchorage, Magic, and others were available but not first-class. The decision was structural — Stripe is a top-tier AWS revenue customer and Coinbase is a major AWS region anchor — and it propagated downstream because hundreds of thousands of net-new AgentCore agents inherit a Privy-or-CDP wallet by default. See Privy x AWS AgentCore default-wallet positioning.

2. Stripe’s Privy acquisition (Jun 2025). Stripe paid a publicly reported $300-500M (some sources cite up to $1.1B inclusive of earn-outs) and folded Privy into the Stripe family alongside Bridge ($1.1B), Tempo, and the USDB stablecoin. From Stripe’s perspective Privy is the layer-4 wallet inside the five-layer Trojan horse. From the embedded-wallet category’s perspective it removed the largest neutral SDK vendor and replaced it with a Stripe-aligned default.

3. Coinbase CDP’s vertical integration with Base + USDC + Onramp. CDP is not just an embedded-wallet SDK — it is a packaged stack of wallet + Onramp + Paymaster + AgentKit + Smart Wallets that routes by default to Base and USDC. This created a parallel default-tier provider with structurally different economics from Privy (CDP earns from on-chain flow; Privy earns SaaS-style fees with Stripe-rail interchange upside).

4. Hyperscaler agent-orchestration platforms (2025-2026). AWS Bedrock AgentCore Payments (2025), Azure AI Foundry Agent Service (2025-Q4), and Google Cloud Vertex AI Agent Builder (2026-Q1) each shipped a managed agent-orchestration surface. Each picked default-wallet providers. The hyperscaler-default war is the immediate channel through which embedded-wallet share is being decided in 2026. Choices visible in public docs: AWS → Privy + CDP; Azure → Privy + (additional default still ambiguous in public docs); GCP → CDP-leaning (Coinbase has structural partnerships, AP2 co-authorship — see AP2 overview).

5. ERC-7715 + ERC-4337 commoditization at the wallet primitives layer. The primitives that embedded wallets sit on top of — ERC-7715 wallet permissions, ERC-4337 account abstraction, ERC-7702 EOA upgrade — became standardized and easier to implement, removing some of the technical moats that smaller SDKs had built around proprietary MPC / TEE choices.

Privy (Stripe)

Owner. Stripe (acquired June 2025; founders Henri Stern + Asta Li remained as Stripe Privy leadership).
Key custody primitive. MPC + TEE shard model. One key shard on the user’s device (Passkey / WebAuthn / Secure Enclave), one shard inside a Stripe-operated TEE (AWS Nitro Enclaves or equivalent).
Default hyperscaler placement. AWS Bedrock AgentCore Payments — co-default with Coinbase CDP. Azure AI Foundry Agent Service lists Privy as a supported default.
Public footprint. ~110M wallets, ~1,000+ developer integrations, including OpenSea, Hyperliquid, Farcaster, Friend.tech, Pump.fun, Story, Zora, and major Stripe Connect merchants.
Strategic role. Layer 4 of the Stripe five-layer collapse. Where Stripe needs a user-identity surface that also acts as the wallet on the Bridge USDB / Tempo rail, Privy is the chosen substrate.

Coinbase CDP

Owner. Coinbase (NASDAQ: COIN).
Key custody primitive. MPC distributed key management, plus a packaged ERC-4337 Smart Wallet integrated with Base Paymaster for gasless USDC transfers.
Default hyperscaler placement. AWS Bedrock AgentCore (co-default with Privy). GCP Vertex AI Agent Builder leans toward CDP (Coinbase is an AP2 co-author with Google).
Public footprint. ~150K developers, ~8K dApps, ~5M monthly active embedded wallets per CDP public communications.
Strategic role. Default-binds Base + USDC; the AgentKit module (2025-04) is the agent-economy SDK; ties the wallet primitive to chain economics that Coinbase already owns.

Magic

Owner. Independent (Tiger / SV Angel-backed).
Key custody primitive. Magic Link (email magic links) plus DKLs MPC 2/2.
Default hyperscaler placement. Not a hyperscaler default in any public AgentCore / Foundry / Vertex doc as of 2026 mid-year.
Public footprint. Long tail of Web3 SaaS clients; pricing has tiered per-MAU SaaS structure.
Strategic role. Pre-Privy market leader, now a secondary-tier provider competing on neutrality (no fintech-conglomerate or exchange parent).

thirdweb

Owner. Independent (Founders Fund, Coinbase Ventures, Polygon — minority stakes).
Key custody primitive. InAppWallet SDK with optional embedded-MPC + smart-account abstraction.
Default hyperscaler placement. Not a hyperscaler default.
Public footprint. Heavy on NFT, gaming, consumer dApps; partial open-source posture.
Strategic role. Niche specialist in consumer / gaming dApps; competes on full-stack tooling (deploy + wallet + payments in one) rather than wallet alone.

Particle Network

Owner. Independent.
Key custody primitive. Modular SDK with MPC + ERC-4337 + own chain-abstraction layer.
Default hyperscaler placement. Not a hyperscaler default.
Public footprint. EVM + BTC + multi-chain coverage; emphasizes a “universal account” model crossing chains.
Strategic role. Differentiates on chain-abstraction depth rather than fintech distribution.

Web3Auth

Owner. Independent (Torus Labs; BNB-Labs investor).
Key custody primitive. tKey + OAuth share reconstruction; open-source core.
Default hyperscaler placement. Not a hyperscaler default.
Public footprint. Long-tail Web3 plus selected enterprise SSO integrations.
Strategic role. Open-source posture is its primary differentiator in a market increasingly dominated by Stripe + Coinbase commercial defaults.

Dynamic

Owner. Independent (a16z crypto-backed).
Key custody primitive. Multi-chain login with embedded + external-wallet bridging.
Default hyperscaler placement. Not a hyperscaler default.
Public footprint. DeFi / consumer-finance dApps that need to bridge embedded + MetaMask users in one session.
Strategic role. Specializes in the embedded-meets-external-wallet UX gap that pure embedded SDKs do not address.

Crossmint

Owner. Independent.
Key custody primitive. Fiat-funded wallet creation; credit-card-to-NFT / RWA receipt rails.
Default hyperscaler placement. Not a hyperscaler default.
Public footprint. NFT, commerce, RWA-tokenization customers; Visa / Mastercard rail integrations.
Strategic role. Specializes in the fiat-funded slice of the embedded-wallet market — the slice that needs card networks rather than crypto rails — and so does not compete head-to-head with Privy / CDP on AI-agent volume.

Hyperscaler default-wallet wars

The most consequential 2026 dynamic is the way hyperscaler agent platforms pre-select wallet providers as defaults inside their managed agent runtimes. Three reasons this matters more than dApp adoption did in 2023-2024:

Provisioning is implicit, not explicit. When a developer spins up an AWS Bedrock agent that needs to receive USDC payments, the AgentCore Payments module provisions a Privy or CDP wallet under the hood. The developer never made a wallet-vendor decision.
Hyperscalers absorb the distribution cost. A secondary-tier SDK now has to convince a developer to opt out of the default — a higher-friction sale than convincing them to opt in to a single chosen SDK.
The default-tier providers earn structural compounding. Every default-provisioned wallet adds to Privy’s or CDP’s installed base, which makes them more attractive as next-default candidates for Azure / GCP and any subsequent hyperscaler-agent platform.

Visible default placements in mid-2026 hyperscaler public docs:

Hyperscaler	Agent platform	Default wallet providers	Notes
AWS	Bedrock AgentCore Payments	Privy + Coinbase CDP	Public since 2025-Q3 GA; x402 + Bazaar MCP integration; see [[agent-economy/x402-http-payment-overview
Azure	AI Foundry Agent Service	Privy listed as supported default; second default ambiguous in public docs	Microsoft’s Stripe relationship is the structural anchor
Google Cloud	Vertex AI Agent Builder	CDP-leaning; AP2 co-authorship with Coinbase is the structural anchor	See [[agent-economy/ap2-overview
Cloudflare	Workers AI Agents	x402-integrated; wallet-provider choice still developer-side	Lower-level: routes payments rather than provisioning identity

Stripe ownership of Privy — strategic implications

Stripe acquired Privy in June 2025. Public reporting has the deal in the $300-500M range; some sources cite up to $1.1B inclusive of earn-out structures. From a category perspective, the consequences:

The largest neutral embedded-wallet vendor is no longer neutral. Non-Stripe merchants and Stripe competitors (Adyen, Block, PayPal) now face a “use my competitor’s wallet stack or roll your own” decision.
Privy becomes Stripe-rail-aware. Inside the five-layer collapse, Privy routes to Bridge (USDB stablecoin) and Tempo (settlement chain) by structural design.
The AWS AgentCore default placement compounds. Stripe-aligned AgentCore agents inherit Stripe-aligned wallets, which inherit Stripe-aligned stablecoin and chain. The vertical stack is monetized at every layer.
Antitrust questions remain open. No US, UK, EU, or Japan competition regulator has publicly intervened in the Privy acquisition as of mid-2026, but the pattern parallels the historical Apple / Google default-search-engine cases (see integrator moat counterpoint section).

M&A activity since 2024

Date	Event	Buyer	Target	Public price
2025-06	Stripe acquires Privy	Stripe	Privy	~$300-500M (some sources up to $1.1B)
2025-06	Stripe acquires Bridge	Stripe	Bridge	~$1.1B
2024-Q3	Coinbase folds Wallet API + Cloud into CDP rebrand	Coinbase	(internal)	n/a
2025-04	Coinbase ships AgentKit on top of CDP	Coinbase	(internal)	n/a
2025-Q3	AWS picks Privy + CDP as AgentCore defaults	AWS / Stripe / Coinbase	(platform default)	n/a
Open	thirdweb / Magic / Web3Auth / Dynamic acquisition rumors	(various)	(various)	None confirmed

Plausibility-weighted but unconfirmed: a hyperscaler may buy a secondary-tier provider to compete with Stripe-Privy and Coinbase-CDP. Public docs and public M&A filings do not show this as of mid-2026.

Sources

Privy public documentation: docs.privy.io
Coinbase Developer Platform public documentation: docs.cdp.coinbase.com
Magic public documentation: magic.link/docs
thirdweb public documentation: portal.thirdweb.com
Particle Network public documentation: docs.particle.network
Web3Auth public documentation: web3auth.io/docs
Dynamic public documentation: docs.dynamic.xyz
Crossmint public documentation: docs.crossmint.com
AWS Bedrock AgentCore Payments public announcement and docs (2025-Q3 GA)
Azure AI Foundry Agent Service public docs (Microsoft Build 2025 / 2026)
Google Cloud Vertex AI Agent Builder public docs (Google Cloud Next 2026)
Stripe public communications on Privy and Bridge acquisitions (June 2025)
FCA, SEC, FSA Japan, and MAS public registers and consultation papers on embedded-wallet and electronic-money classification