$ZKP Core question: As the window for technical leadership narrows and the token’s valuation falls 70% from its Series A level, where is zkPass most likely—and where should it—go in 2026–2028?
Main thesis: The focus will shift from “proving Web2 data to humans” to “providing AI Agents with verifiable trust and credit”; institutional compliance is the only second revenue stream capable of generating real income within 24 months; node permissionlessness determines the valuation floor.
Technical catch-up: TLS 1.3 coverage, plugin-free mobile proofs, proof aggregation, and IP-block-resistant pools are prerequisites—not optional extras—for entering high-frequency Agent scenarios.
Business model: shift from "charging per proof" to "enterprise subscriptions plus revenue sharing on credit credentials"; purely crypto-native demand cannot support the current cost structure.
Biggest risk: the Agent narrative is running ahead of real call volume; monthly unlocks of approximately 1.66% of total supply are a hard constraint.
Starting assessment: all three curves have hit their ceiling simultaneously; transformation is no longer optional
zkPass's growth over the past three years was built on three assumptions: technological leadership, growing demand to put Web2 data on-chain, and the token's ability to capitalize on these capabilities. By the end of 2026, all three assumptions have weakened at once—competitors have closed the technology gap, human-side demand is constrained by a small number of major customers, and token prices are driven by unlocks and leverage. This means incremental growth can only come from new use cases; improving the existing business is no longer enough.
The window for technological leadership has closed; differentiation must shift from "faster" to "more trustworthy"
By 2026, the zkTLS sector has moved from "who can build it" to "who can be trusted and adopted at scale." Reclaim Protocol has achieved the largest production footprint with its Proxy model: over 1 million verifications, 25 customers, 30 chains, nearly 900 community data sources, and plugin-free mobile proofs in 2–4 seconds. Primus (formerly PADO) added TEE-based attestation of witness nodes themselves on top of the base protocol, addressing the core concern that "witnesses might act maliciously." zkPass originally differentiated itself through a hybrid model (Proxy as the primary method, with MPC as a fallback) and sub-second proofs enabled by VOLEitH. But as "fast" is no longer a unique advantage, the comparable dimensions have shifted to witness trustworthiness, institutional acceptance, and ecosystem standards.
There is a structural weakness that must be taken seriously: Proxy nodes have fixed IPs and behavioral patterns, making them identifiable and blockable by target websites. If a major bank or government portal chooses to block zkTLS proxy traffic, that data source becomes unavailable to all users. zkPass's MPC fallback mitigates this, but MPC has very high bandwidth costs—a 100 KB data exchange may require the user to upload about 32 MB. So, in engineering terms, "censorship resistance" is still conditional.
The ceiling for human-side zkKYC is determined jointly by customer concentration and countermeasures by data sources
Privacy-preserving identity proof for individuals is zkPass's earliest and most fully developed narrative: selective disclosure, locally generated proofs, and reusable proofs. But paying customers for this line are highly concentrated—exchanges, brokerages, payment institutions, and RWA issuers are all limited in number, have long decision-making chains, and face heavy compliance scrutiny. For them, adopting zkTLS instead of "uploading documents" is not purely a technical choice; it partially shifts KYC responsibility from "storing data themselves" to "trusting an external proof system." This requires legal opinions, regulatory engagement, and carefully defined liability boundaries. This makes the human-side market stable but slow-moving, and unlikely to support the high-growth expectations of an infrastructure project.
The token has entered a supply-led phase; the valuation anchor has shifted from narrative to liquidity
As of November 20, 2026, ZKP is quoted at approximately $0.0305, with a market cap of about $6.14 million, an FDV of about $30.47 million, and a circulating supply of 201.7 million tokens (20.2% of total supply), down about 87% from its all-time high of $0.23. Compared with the $100 million token FDV in the 2024 Series A SAFT, the current price is more than 70% below the institutional entry valuation. The market structure is even more telling: perpetual open interest is about $15 million, or 244% of market cap, and derivatives trading volume is 3.4 times spot volume—prices are being set by leverage, not fundamentals. Meanwhile, according to third-party unlock data, around 16.62 million tokens (about 1.66% of total supply) enter circulation around the 19th of each month.
Curve
State at the end of 2026
Inflection-point signal
Strategic implications
Technology
Hybrid zkTLS remains a differentiated design, but Reclaim and Primus have caught up with or surpassed it in individual areas such as speed, cost, and ecosystem scale
A competitor secures its first production contract on the institutional side
The basis of competition must shift to "auditable and acceptable to institutions"
Demand
Human-side customers are concentrated among exchanges and licensed institutions, with integrations taking quarters; countermeasures by data sources could reduce availability at any time
A single major customer accounts for most call volume
Go deep on the human side, not broad; shift focus to new sources of demand
Capital
FDV of approximately $30 million, 20.2% circulating, leverage/market-cap ratio of 244%, ongoing monthly unlocks
Net buying has absorbed unlocks for two consecutive months
Token policy needs to be tied to real revenue; otherwise, every rally is a selling opportunity
Thesis One: Build a robust Agent trust layer—the only direction with room to grow by an order of magnitude
With all three curves hitting their ceiling, the most promising incremental opportunity isn't "more human users proving more Web2 data" but an entirely new source of demand: autonomous AI Agents. Agents need to place orders, make payments, borrow, and receive authorization, but the market has no mechanism to determine "who operates the Agent behind the scenes, and whether they're creditworthy." This is a problem that falls squarely within zkTLS's capabilities.
Real demand exists: the Agent economy lacks a layer that ensures "data provenance is trustworthy"
In 2026, machine payments and Agent authorization protocols (such as x402-style pay-per-call and AP2-style delegated authorization) are spreading rapidly, and Agents are already starting to perform actions on people's behalf that have economic consequences. The accompanying risk-control and credit layers are also taking shape—t54's Trustline/Claw Credit scores Agent transaction risk and provides credit, while KEEL builds bonds and credit lines for Agents on Solana. But all these solutions share a missing foundation: is the data they use for assessment genuine and from an authoritative source? On-chain activity can be faked, self-reported performance can be fabricated, and centralized KYC covers only the "person," not the "Agent's operator." zkPass's niche is exactly here: it doesn't build risk models; it supplies "trusted data sources."
This is also why the website changed its positioning in 2026 to "Proof for humans. Credit for agents."—one cryptographic toolkit for two economies. Technically, it holds together: exchange balances, trading histories, platform reputations, and task completion records all sit behind HTTPS, while the TransGate SDK-B already offers three calling modes—verify(), verifyCampaign(), and chat()—with chat() able to parse natural-language intent directly into a proof.
Productizing the offering in three layers: operator endorsement, behavioral track record, and credit credentials
Making "issuing credit to Agents" a reality requires breaking it down into three deliverable layers, rather than stopping at positioning copy. The table below sets out what each layer proves, which existing capabilities can be reused, and criteria for determining success or failure within 12 months.
Layer
What needs to be proven
Existing capabilities to reuse
Success criteria within 12 months
Identity layer
There is a real, non-Sybil operator with financial capacity behind the Agent: exchange account balance, account age, KYC status
TransGate extension + balance-related schemas
At least 3 Agent platforms make "operator endorsement" a prerequisite for access
Behavior layer
The Agent's historical track record: transaction records, task completion rate, platform reputation, accumulated API calls
Multi-source bundling with verifyCampaign, Schema Market templates
A cross-platform, reusable "Agent track record" standard emerges, not exclusive to zkPass
Credit layer
Package the track record into credentials that lending/risk-control protocols can consume directly, replacing overcollateralization
On-chain attestations, composable proofs
At least 1 lending or credit protocol uses its credentials to adjust credit limits or interest rates
The commercial value of the three layers is uneven: the identity layer is easiest to implement because demand-side players (Agent platforms and exchanges) already have anti-Sybil budgets. The credit layer is the most valuable, but depends on risk teams adapting their models and has the longest timeline. The sensible sequence is to start with the identity layer to secure integrations, then use those integrations to gain influence over the credit layer.
Set the success criteria up front: without default integration by more than three platforms, this is still just a PowerPoint deck
The biggest risk in the Agent narrative is that "the story gets ahead of reality." Set falsifiable thresholds internally: within 12 months, at least 3 agent-native payment or credit protocols must use zkPass as a default attestation source, SDK monthly call volume must reach around 100,000, and there must be paying customers—not just subsidies. Falling short means this is merely a market narrative, and resources should be redirected to institutional compliance.
Competition also needs to be assessed clearly: centralized KYC providers are rolling out "Agent operator verification," on-chain analytics firms are selling "Agent scoring," and TEE solutions are selling "hardware-based trust." zkPass's relative advantage is the authenticity of its data sources (raw data from banks, exchanges, and platforms); its relative disadvantage is the latency and cost of each proof—which leads directly to the four technical shortcomings in the next section.
Thesis Two: Institutional compliance is the only path to real revenue within 24 months
The Agent thesis determines the upside; the institutional thesis determines whether the project can survive. zkKYC/KYB has clearly defined paying customers, existing budgets, and quantifiable pain points, making it the most realistic source of revenue today.
The value proposition can be translated into liability and cost savings, not just "more privacy"
For exchanges and payment institutions, zkTLS's real selling point isn't user privacy, but a smaller liability footprint: turn "storing 1 million passports and bank statements" into "storing 1 million zero-knowledge proofs." The data at risk in a breach disappears, and the way to provide evidence in response to regulatory inquiries shifts from "what did we store?" to "what did we verify?" This logic is more readily accepted in jurisdictions with stricter data-protection enforcement (the EU, UK, and some Asia-Pacific markets), and is easier to justify in procurement.
Moving from "selling a tool" to "taking on liability" requires three pieces to be put in place
First, legal and audit endorsement: data-minimization design, explanations of cross-border data flows, and regulatory engagement on whether "proofs can serve as KYC records" are not engineering problems. Second, channel positioning: rather than directly replacing established KYC providers such as Sumsub and Onfido, it is more feasible to serve as a pre-filter for them—use proofs to screen out most low-risk users first, sending only high-risk cases through the full KYC process. This minimizes channel resistance and makes it easier to enter existing customer accounts. Third, enterprise delivery capabilities: SLAs, white-labeling, audit logs, and incident-response procedures are hard requirements for procurement by licensed institutions, yet publicly available materials still focus mainly on protocol capabilities.
The failure condition is clear: no production contracts with licensed institutions within 6–9 months
The compliance thesis cannot be measured by "pilots." If 2–3 licensed institutions have not signed production contracts within 6–9 months (not PoCs or joint research), it can largely be concluded that this path cannot scale under the current regulatory environment. Resources should then be redirected to the Agent thesis or the node network.
Thesis Three: Node permissionlessness sets the valuation floor; without it, the token is just a governance vote
For the token to become a "settlement asset," the network must perform real verification work that needs to be secured by staking. This step has not yet been completed.
Current state: whitelisted Beta and a single verifier are the biggest holes in the "trustworthy" narrative
The node network has a complete economic structure designed: Pool Nodes (Validators and TEE Nodes), Watcher monitoring and arbitration, staking thresholds, suspension and ban lifecycle, and unstaking cooldown periods. But the network is currently explicitly in a whitelisted Beta access phase, on the grounds of "maintaining verification quality until governance matures." Independent research pointed to an even sharper problem back in 2025: the verifier implementation and extension were closed-source, only the project team could run nodes, and users had to trust that it would not deny service or collude with the prover. Subsequent SDK open-sourcing and the node Beta have eased some concerns, but permissionless access is still not in place.
Three steps: permissionless access, objectively decidable slashing, and TEE node attestation
The first step is to open up access, supported by tiered staking and reputation—permissionless access inevitably brings in low-quality nodes, so they should be filtered by economic weight rather than administrative review. Second, make slashing conditions objectively decidable: many current disputes concern whether a proof failed because "the data source changed its site" or because "the node acted maliciously." Subjective judgments turn slashing into a governance battle. A workable approach is to build a reproducible library of failure cases and turn disputes into replayable test cases. Third, add TEE remote attestation to nodes, turning "the witness itself is trustworthy" from an economic assumption into a hardware-based one. This is precisely the layer Primus adds on top of the base protocol, and the point institutional clients are most likely to ask about.
The value of these three steps isn't technical elegance, but token economics: settlement fees and staking demand only make sense when verification work is real, competitive, and slashable. Otherwise, ZKP is just a governance vote.
Technical roadmap: four shortcomings that must be addressed
For the human use case, the following four items are "optimizations"; for the Agent use case, they are "the price of entry." The reason is the difference in call frequency: an Agent may request proofs hundreds of times more often per day than a person. Per-proof cost and latency must fall by one or two orders of magnitude, or the business model won't work.
Shortcoming
Current state
2027 target
Success criteria
TLS 1.3 support
TLS 1.2 is mature; TLS 1.3 support is still evolving across zkTLS solutions
Full TLS 1.3 compatibility with mainstream data sources, without significantly increasing proof time
TLS 1.3 sites account for more than half of the available data sources
Plugin-free mobile
The human use case depends on a Chrome extension; the mobile experience lags behind Reclaim's plugin-free 2–4-second proofs
Generate proofs directly in mobile browsers, with no installation required
Mobile proofs account for more than 30%
Proof aggregation and recursion
Each proof is settled independently, and on-chain costs and verification overhead grow linearly with call volume
Aggregate proofs in batches, reducing the cost per settlement to a level that supports high-frequency calls
Cost per proof in batch scenarios falls by an order of magnitude
Censorship-resistant transmission
Proxy node IPs can be identified and blocked; the MPC fallback has high bandwidth costs
Run multiple nodes in parallel and rotate proxy pools to keep key data sources available
High proof success rates for the top 20 data sources
Of these four items, "plugin-free mobile" is the easiest to underestimate. The Agent use case may sound unrelated to browsers, but Agents are operated by people, and the experience of authorizing and endorsing from a phone directly determines this system's conversion rate.
Token economics: shift from narrative premium to settlement demand
ZKP's economic design itself has no obvious flaws: a fixed supply of 1 billion with no further issuance, partial burning of settlement fees, periodic DAO buybacks, staking required for validators, and no team or investor unlocks at TGE. The problem lies on the demand side—if proofs are mainly driven by subsidies and airdrops, "settlement fees" are just circular flows on paper.
Three figures define the current constraints
The circulating ratio is 20.2%, monthly unlocks amount to about 1.66% of total supply, and perpetual open interest is 244% of market cap. Together, these mean that any narrative-driven rally will be quickly absorbed by unlocks and leveraged position liquidations. In April 2026, after the team's multisig wallet transferred 10 million ZKP (about 3.7% of the circulating supply at the time) to an exchange, the price fell 7.35% that day—an observable example.
Improvement priorities: make fees real, tie rewards to tasks, and disclose unlocks transparently
First, make "who pays for proofs" visible in a public ledger: track calls separately for enterprise customers, protocol customers, and subsidized users, so airdrop claims aren't counted as demand. Second, tie validator rewards to real task volume, rather than relying on inflationary subsidies to maintain incentives; otherwise, staking size is merely a mirror of inflation. Third, although the unlock schedule is fixed by contract, treasury policy can be more transparent—regularly disclosing the use of funds and linking buybacks to revenue milestones can reduce expectations of "unconditional selling pressure."
Scenario
Conditions for realization
Business indicators
Valuation range (FDV)
Bullish
Agent payments become mainstream, zkPass becomes the default attestation source, and more than 3 licensed institutions are onboarded
Paid calls scale steadily, and staking grows in line with real task volume
$80 million–$120 million, approaching Series A valuation
Base case
The Agent thesis remains at the pilot and small-scale integration stage; the compliance thesis advances slowly; nodes remain partially permissioned
Call volume grows moderately, and revenue barely covers operating costs
$30 million–$60 million, fluctuating around current levels
Bearish
The Agent narrative falls through, institutional contracts fail to materialize, and unlocks continue
Call volume stagnates, and the token is primarily a governance and speculative asset
Below $15 million
The ranges above are conditional scenarios, not price predictions. They are driven by a single variable—"whether verifiable paid demand emerges"—rather than by technological superiority. That's why the milestones in the next section must focus on payments and integrations, not feature releases.
Roadmap: three phases and verifiable milestones
Phase
Timeline
Core objective
Verifiable milestones
Phase One: Validate Agent commercialization
Q4 2026–Q2 2027
Turn SDK-B from a capability into revenue: first implement operator endorsement, while advancing the behavioral track record standard
Default integration by more than 3 Agent platforms; public ledger of paid calls; meet mobile and TLS 1.3 coverage targets
Phase Two: Advance compliance and permissionless access in parallel
2027
Sign production contracts on the institutional side; move the node network from whitelisting to permissionless access and make slashing conditions objective
Contracts with 2–3 licensed institutions; external nodes make up more than half the network; at least one slashing dispute adjudicated using replayable cases
Phase Three: Settlement standards and cross-chain recognition
2028
Turn schemas and credit credentials into cross-protocol standards, with settlement fees becoming the primary revenue source
External protocols cite its schemas as de facto standards; settlement fee revenue exceeds subsidy-driven call revenue
Conclusion: the upside depends on Agents, the floor on nodes, and cash flow on compliance
First, the Agent trust layer is the right direction, but it must be tested against integration numbers and paid calls. It addresses a genuine gap in the Agent economy—without trusted data sources, any credit model is built on sand—and elevates zkPass from an "identity tool" to "infrastructure." But so far this thesis has only positioning and an SDK, with no evidence of third-party adoption. If it fails to meet the thresholds within 12 months, it should be scaled back.
Second, institutional compliance determines short-term survival, and the path should be a "pre-filtering layer," not a "replacement for KYC providers." Translating the privacy narrative into liability and cost savings is what can get the product into procurement processes. Failure to secure 2–3 production contracts would show that this market is not viable under the current regulatory environment.
Third, node permissionlessness is the valuation floor, not the ceiling. Without permissionless access, slashability, and provable verification work, the token's settlement utility is empty. Conversely, if this step is completed, the token will have a value anchor tied to call volume even if business growth is slow.
Scope and limitations. Business scale figures cited in this article follow official figures (such as 200M TVC, 300+ schemas, and 80+ integrators). TVC is a cumulative count of computations, not unique verifications, and cannot be directly compared with competitors' "verification counts." Circulating supply figures vary by source (CoinGecko lists 201.7 million; on-chain reporting cited 268 million in April 2026). Using the latter figure would raise the market cap estimate by about 30%. In addition, several projects share the ZKP ticker; one presale project named in an on-chain investigation is unrelated to zkPass, so contract addresses must be checked when citing trading data.
Monitoring checklist (in priority order):
Number of third-party integrations and monthly call volume for the Agent SDK; whether paying customers have emerged
Number of production contracts on the institutional side, not pilots or joint research projects
Share of external nodes in the node network, and whether the first objective slashing has been completed
Proof success rates for major data sources, as a proxy measure of censorship resistance
Actual destination of monthly unlocks: treasury, team, or exchanges
Ratio of perpetual open interest to market cap, to determine whether prices are driven by leverage or demand
Decision guidance. For investors, this is an option-like asset whose technology has been validated but whose business model has not. The valuation is not expensive, but the supply structure is unfavorable; position sizing should follow milestones, not the narrative. For ecosystem developers, the cost of integrating now is lowest and negotiating leverage is greatest, but prioritize use cases with paying customers (exchange access, Agent platform risk controls) over pure airdrop opportunities. For the project team, the recommended priority is to secure cash through compliance, integrations through Agents, and trust through nodes; only completing all three will finish this leg of the journey.$ZKP

