Distilled from the 2026-05 landscape sweep + 25 PMM reviews. Each is a structural shift in the cohort, not a single-competitor observation.
Stripe/Privy (Jun '25), Consensys/Web3Auth (Jun '25), Fireblocks/Dynamic (Oct '25), Polygon/Sequence + Coinme combined ~$250M (Jan '26). Independent embedded-wallet category is consolidating into payments + custody platforms. WaaP's "independent, not absorbed" positioning is the explicit counter.
Five named protocols (ERC-8004, x402, MPP, AP2, ACP) now define authorization model + identity layer + payment rail. Company comparison alone misses the picture — WaaP needs protocol-level interop framing.
Turnkey shipped Human-Agent Consensus (cryptographic co-approval) early 2026. Privy has m-of-n quorum. WaaP's unique angle is now conversational HITL (Telegram + email one-tap), not "we have HITL at all."
Lit Protocol + NEAR Chain Signatures both shipped decentralized MPC at scale before Ika ramped (NEAR: 2.5M Sweat Wallet users; Lit: V3 on Phala). Don't claim "first decentralized wallet infrastructure." Lead with non-collusive-by-cryptography + identity-bound + compliance-aware.
x402 hit 165M+ tx, ~$50M volume, 69K active agents by late Apr 2026. AWS Bedrock AgentCore Payments (May 7, 2026) named Coinbase + Stripe + Privy as wallet partners. WaaP doesn't support x402 yet + no Bedrock channel — both are real distribution gaps.
All 25 competitors by tier × segment. Direct = full profile, named threat. Emerging = full profile, smaller scale. Stage 2 = one-line index entries pending full profile.
| Apps lens | Agents lens | Dual-use (apps + agents) | |
|---|---|---|---|
| Direct(25) | CDP Embedded Wallets Cometh Connect Openfort Para Sequence (Polygon) | Coinbase Agentic Wallets Skyfire Catena Labs Halliday MoonPay Agents ChaosChain NEAR Chain Signatures Cobo Agentic Wallet | Privy (Stripe) Turnkey Crossmint Web3Auth (Consensys) Lit Protocol Magic Labs Dynamic (Fireblocks) Particle Network Tether WDK |
| Emerging(3) | — | AgentaOS Sponge Orthogonal | — |
| Stage 2 stubs(11) | Thirdweb Alchemy Dfns Portal HQ | Bankr (on Privy) HeyElsa Coral Protocol Synergetics.ai Giza ARMA Brahma Ethy AI Almanak Wayfinder Theoriq | — |
Where each cohort sits on the custody axis. The "key-share split, never reconstructed" property is unique to WaaP / Ika; everyone else concentrates the whole key in one trust domain.
| Custody architecture | Players | Trust assumption |
|---|---|---|
| CustodialServer-held keys, API-key access | Sponge Skyfire | Full trust in operator infrastructure |
| Single-TEEWhole key in one enclave | Privy Coinbase Agentic Wallets Magic Labs Turnkey Lit Protocol V3 CDP Cometh Connect Crossmint Openfort TEE | TEE attestation; TEE compromise = full exposure |
| TSS-MPC / Thresholdn-of-m share split, key never reconstructed (variant approaches) | Dynamic (TSS-MPC) Particle Network (2-of-2) Web3Auth (SSS / MPC-TSS) Para (DKG) | Threshold of nodes / shares can in principle collude |
| Decentralized MPCMulti-node MPC network | NEAR Chain Signatures Lit Protocol Naga (sunset Apr '26) | Network threshold can in principle collude (NEAR: 8 validators + EigenLayer) |
| Self-custodial toolkitIntegrator builds the custody primitive | Tether WDK MoonPay Agents (OWS) | Integrator-chosen; no single primitive enforced |
| Hybrid 2-of-3Agent + server + user; any 2 can sign | AgentaOS | Agent + server can sign without user; user share is recovery only |
| 2PC user-shareUser device + secure enclave; BOTH required every signature | WaaP | Non-collusive by cryptography — user share required every signature; network compromise survivable |
Note: Turnkey shipped Client-Side Signing (opt-in, Apr 30 2026) — whole key on device, still single-trust-domain. Not 2PC.
Five named protocols define the agent payments + identity layer. WaaP's relationship to each: compete / compose / gap / adjacent.
ERC-8004
MetaMask + EF + Google + Coinbase co-authors → Ethereum standard
Competes (identity)Mainnet Jan 29 2026. Identity / Reputation / Validation registries on-chain. Emerging standards-level competition to Human Passport's agent identity surface — different design philosophy (registries vs pluralistic Stamps), not direct substitute. ChaosChain is reference implementation.
x402
Coinbase → Linux Foundation x402 Foundation (Apr 2026)
Gap (no support)HTTP-native pay-per-request. 165M+ tx / ~$50M volume / 69K active agents by late Apr '26. 22-org foundation incl. AWS, Stripe, Visa, Mastercard, Google, Microsoft, Solana Foundation. WaaP doesn't support x402 yet. Closest standard to WaaP's on-chain scenarios. Gap is widening.
MPP (Machine Payments Protocol)
Stripe + Paradigm via Tempo L1
Adjacent (different layer)Tempo mainnet Mar 18 2026. Series A $500M @ $5B valuation. "OAuth for money" session primitive. ISO 20022-aligned per Tempo positioning. Lightspark extends to Bitcoin Lightning. Different authorization model from WaaP Privileges (session vs key-share-bound).
AP2 (Agent Payments Protocol)
Google → FIDO Alliance (donated Apr 28 2026)
Composes (via x402 extension)60+ partner organizations per Google's donation. v0.2 "Human Not Present" payments. Mandates as W3C VCs. A2A x402 extension co-developed with Coinbase + EF + MetaMask. Verifiable Intent co-developed with Mastercard.
ACP (Agentic Commerce Protocol)
Stripe + OpenAI
Adjacent (Web2 commerce)ChatGPT Instant Checkout shuttered Mar 4 2026 (~30 merchants live; scaling bottleneck on AI side). ACP continues for app-based commerce (Instacart, Target) + Microsoft Copilot Checkout + Salesforce-Stripe. Web2-commerce framing, mostly card rails. Less relevant to WaaP's on-chain scenarios.
Three architectural claims that survive technical scrutiny. Each is most material against a different competitor cohort.
Non-collusive by cryptography
2PC-MPC requires the user share for every signature. Network nodes can be fully compromised; assets remain secure as long as user share is uncompromised.
Identity-bound
Human Passport — pluralistic proof-of-personhood (175+ campaigns served) as input to agent authorization. Verified-human agent ownership.
Compliance-aware
zkKYC + Proof of Clean Hands. Privacy-preserving compliance primitives. Regulatory positioning peers don't have.
M&A activity in the embedded-wallet / agent-wallet cohort. The "independent, not absorbed" positioning wedge sharpens each quarter.
Counter-signal: Stripe (founders' alma mater) chose Coinbase, not Sponge (3-person ex-Stripe team), for AWS Bedrock AgentCore Payments — the natural-acquisition thesis for Sponge is materially weakened.
"We have HITL" is no longer a differentiator. The framing now is how the human approves — conversational (Telegram, email), cryptographic (key signature), programmatic (rule), or none.
| Mode | Approval mechanism | Players | Best-fit operator |
|---|---|---|---|
| Conversational | Telegram + email one-tap approval; user already has these channels open | WaaP | Individual agent operator supervising AI agent activity in real time |
| Cryptographic co-approval | Co-approver holds a key and signs; m-of-n quorum | Turnkey (Human-Agent Consensus) Privy (m-of-n quorum) | Enterprise risk team / treasury operator already on cryptographic-quorum workflows |
| Programmatic only | Pre-defined policy / guardrails enforced at signing; no real-time human | Lit (Vincent Policies) Halliday (workflow guardrails) Cobo (Pact-bound) Sequence (Smart Sessions) | Autonomous workflows where pre-defined scope is enough |
| No HITL primitive | Agents operate within API-key boundary; no approval flow | Coinbase Agentic Wallets Skyfire Sponge AgentaOS Orthogonal MoonPay (OWS spec) Tether WDK Magic Newton Particle Network | Fully-autonomous-by-default builders |
Things competitors have that WaaP doesn't. Knowing these in advance prevents being caught off-guard in a sales conversation.
x402 protocol support
No native x402 client. Coinbase + Dynamic + Lit ship it; Crossmint integrates; 165M+ tx ecosystem.
Most material for agent payment flows. Roadmap priority worth elevating.
AWS Bedrock AgentCore Payments
Coinbase + Stripe + Privy named as wallet partners in May 7 preview. WaaP not in this enterprise channel.
Hyperscaler distribution moat that's hard to bypass without channel-level partnership.
MiCA passport (EU regulatory)
Crossmint (CNMV Jan '26) + Cometh Connect have MiCA. WaaP cannot serve regulated EU fintech sales without it.
Lightly contested segment — only two players have it. Real moat for EU enterprise wins.
Banking-rails partnership at scale
Turnkey × Lightspark Grid (65+ banking rails + Visa debit) + Dynamic via Fireblocks. WaaP has no equivalent.
Closes the on-chain → bank-rails loop for institutional buyers.
Gaming customer roster
Sequence (Ubisoft, Take-Two, Immutable, Magic Eden) + Magic (Mattel, Forbes, Polymarket) + Halliday (DeFi Kingdoms, Shrapnel) + Thirdweb (Unity/Unreal SDKs).
Most distinct + durable segment in the cohort. WaaP has no gaming logos.
Stablecoin volume at scale
Turnkey $100B+ since 2023 (self-reported). Tether WDK has USDT distribution. WaaP doesn't surface stablecoin-volume figures.
Affects positioning against stablecoin-payments-first builders.
Coinbase ecosystem stack
CDP Embedded + CDP Server + Agentic Wallets + AgentKit + x402 + Base + USDC Rewards. Each Coinbase product reinforces the others.
"Compete with infrastructure" problem. Counter-positioning: independent, not stacked.
Standards-level "founding member" status
x402 Foundation (22-org Linux Foundation), AP2 FIDO partners (60+), OWS co-signatories (20). WaaP isn't on these signatory lists.
Worth pursuing standards-cohort membership where it aligns with WaaP's stack.
Grounded Q&A over the full competitor corpus. Answers cite source profile paths. Token required (1Password → "WaaP Research Bot Tokens"). Verify any figure / date / partnership before external citation.