# Grass

> Bandwidth-sharing DePIN network that claims 8.5M users. How to earn GRASS, tokenomics, and why 190-country scale doesn't mean decentralised.

- Scores: Freedom 3.7/10 (grade F), Returns 5.5/10 (grade C)
- Category: data · Token: GRASS
- Published: 2026-03-03 · Updated: 2026-10-09
- Canonical: https://ownyourmind.ai/projects/grass/
- Data (scores + integrity verdicts, JSON): https://ownyourmind.ai/api/projects/grass.json
- Cite as: Own Your Mind (ownyourmind.ai), licence CC BY 4.0. Link the canonical URL.

Charts and interactive widgets are omitted here; their data is in the JSON above.

---

## What it does

Grass is a bandwidth-sharing network on Solana where users install a browser extension, desktop app, Android app, or the Grasshopper hardware device to contribute unused internet bandwidth to scrape public web data for AI training. The pitch is simple: you have spare bandwidth, AI companies need web data, Grass connects the two and you earn tokens. Grass claims over 8.5 million users across 190 countries, and reported retrieving 3 petabytes of data a day in September 2025.

The architecture has five layers. Grass Nodes are lightweight clients that relay web traffic using residential IP addresses, available as browser extensions, desktop apps, mobile apps, and since Q4 2025 the Grasshopper hardware device, a palm-sized aluminium box that plugs into your router via Ethernet and runs 24/7 without touching your personal devices. Routers are intermediary infrastructure connecting nodes to validators and managing network accountability. Validators process and verify web transactions, generating zero-knowledge proofs to checkpoint session data on-chain. A ZK Processor batches these proofs and submits them to Solana. Edge Embedding Models convert the raw scraped data into structured formats suitable for AI training. A Sovereign Data Rollup on Solana is planned to provide verified data with full provenance tracking, though its operational status remains unclear despite being described as "launching soon" throughout 2024 and 2025.

Live Context Retrieval (LCR) is the enterprise product. When an AI model receives a prompt, Grass retrieves relevant current information from the web to supplement the prompt before it reaches the model. Every LCR call requires GRASS tokens; fiat payments are converted to GRASS.

Grass's 28 September 2026 post, "Abundant Intelligence", sets out a shift toward retrieval at inference time. It says a Contents API will let a model open and render any public webpage through the network's residential connections, and that a Search API will follow once Grass has indexed enough of the public web, with multimodal grounding alongside. Grass opened the Contents API on 8 October 2026 behind an access request, adding JavaScript rendering, geo-targeting and freshness controls (Grass on X). The Search API and multimodal grounding are still described as coming.

The project was founded by Andrej Radonjic (CEO) and Chris Nguyen (CTO). Radonjic holds a Master's in Mathematics and Statistics from York University and a Bachelor's in Engineering Physics from McMaster. He was pursuing a PhD in computational physics before founding Wynd Labs during the pandemic. Nguyen is identified through a Backblaze case study and Wellfound but maintains a limited public profile. Radonjic is the most publicly identified team member. The operating entities are Wynd Labs (headquartered in Toronto, Canada; incorporation jurisdiction not publicly disclosed) and the Grass Foundation (a separate entity managing token contracts and revenue). Team size is undisclosed.

The GRASS token launched on 28 October 2024 via Airdrop One, distributing 100 million tokens (10% of supply) to 2.8 million eligible wallets. No ICO or IDO was held.

## Value proposition

AI companies need vast quantities of web data, and millions of residential internet connections can provide it at scale, with better coverage and lower detection rates than data centre scraping. Residential IPs bypass bot detection and geo-restrictions that block data centre connections, giving Grass a structural advantage in data collection. This is factually accurate.

Grass's homepage claims over 8.5 million users (October 2026), and its last network update on X, in September 2025, reported 8.3 million users running a node in the prior 30 days, 1 million concurrent connections and 3 petabytes retrieved daily. Approximately 20 enterprise customers, described as including "the largest AI labs," with "seven-figure revenue clients" (Grass's own characterisation). The VALID dataset, a collaboration with LAION and Ontocord, produced 14 terabytes of multimodal data from 720,000 CC-BY YouTube videos (30 million audio snippets, 15 million images). ClipTagger-12b, a 12-billion parameter vision-language model built with Inference.net, reportedly benchmarks favourably against GPT-4.1 and Claude 4 Sonnet at 15x lower cost. These are tangible outputs.

The airdrop was exceptionally wide. 2.8 million recipients received 10% of supply with no lockup. 88.36% was claimed by 2.06 million wallets. For context, most Solana airdrops reach a few hundred thousand addresses. Grass reached 2.8 million. The minimum threshold of 500 points (roughly 100 hours of uptime) was accessible to anyone with an internet connection.

Counter-narrative: "Decentralised data layer for AI" is the tagline, but the data layer is centralised. Wynd Labs and the Grass Foundation control all enterprise relationships, data storage (centralised on Backblaze B2, 2-3 petabytes uploaded daily), validation (transitioning from a single centralised validator), and revenue (customer contracts run through Grass DataCo, a Foundation subsidiary, and the only published figures come from one commissioned review). Users contribute bandwidth but have zero control over what is scraped, who buys the data, pricing, or protocol direction. Bandwidth rental, not data ownership.

Closed-source is the hardest problem. Wynd-Network's GitHub organisation has zero public repositories. GrassFoundation has one repository (GrassDownload, an HTML page). Node software, router software, validator software, and the ZK processor are all proprietary. You can't independently verify claims about zero-knowledge proofs, data handling, or network architecture.

## Tokenomics

GRASS is an SPL token on Solana (contract: Grass7B4RdKfBCjTKgSqnXkqjwiGvQyFbuSCUJr3XXjs). Total supply: 1 billion, with mint authority still held by a Squads multisig, so the cap is policy rather than code. Circulating supply has been growing through 2025-2026 as vesting unlocks occur, with a significant 55 million token unlock to contributors on 28 February 2026. See live data above for current circulating figures.

Distribution:

- **Early Investors:** 25.2% (252M GRASS; 1-year cliff reached October 2025, then linear over 12 months; approximately 39% unlocked as of March 2026)
- **Foundation and Ecosystem Growth:** 22.8% (228M GRASS; cliff period then linear over 60 months; managed by Grass Foundation; approximately 37% unlocked)
- **Contributors (Team):** 22% (220M GRASS; 1-year cliff reached October 2025, then linear over 36 months; approximately 14% unlocked)
- **Future Incentives:** 17% (170M GRASS; linear over 36 months, no cliff; originally allocated toward Season 2 incentives. Approximately 68% has unlocked on the emissions schedule. The Grass-native wallet launched on 23 July 2026 and Season 2 claims went live the same day, but those rewards are paid in USDC, not new GRASS emissions)
- **Airdrop One:** 10% (100M GRASS; 100% unlocked at TGE; distributed to 2.8M wallets; 88.36% claimed)
- **Router Rewards:** 3% (30M GRASS; linear over 30 months; approximately 55% unlocked)

The insider allocation is 47.2% (25.2% investors plus 22% team). The Foundation controls an additional 22.8%. Community allocation is 30% (10% Airdrop One plus 17% Future Incentives plus 3% Router Rewards). The October 2025 cliff released approximately 181 million investor tokens (50% of the investor allocation). A further 55 million tokens unlocked to contributors on 28 February 2026, representing approximately 13% of circulating supply at the time. Monthly linear unlocks continue across all categories through 2028-2029.

No burn mechanism exists. Staking yields approximately 7.49% APR via delegation to routers per Stakin, a staking provider, with a 7-day unstaking period. Slashing is planned but not currently implemented; when it arrives, it will be manual and multisig-controlled.

Funding totals $14.5 million across four rounds: $1 million pre-seed (No Limit Holdings, July 2023), $3.5 million seed (Polychain Capital and Tribe Capital co-leads, December 2023), undisclosed Series A (HackVC lead, with Delphi, Brevan Howard Digital, Lattice, September 2024), and $10 million bridge/token purchase (Polychain and Tribe Capital, October 2025). Polychain Capital invested three times. On 29 September 2026 Multicoin Capital announced an investment in Grass from its hedge fund and venture fund, with no amount disclosed.

GRASS is spot-listed on Bybit, Gate.io, KuCoin, Kraken, Bitget, Bithumb, OKX (added 24 April 2026) and Coinbase (spot trading from 26 August 2026), with Binance Futures perpetual contracts (up to 75x leverage) adding derivatives liquidity. The ATH of $3.89 was hit on 8 November 2024, just eleven days after launch. The token was about 89% below that peak on 24 September 2026 (CoinGecko).  See live data above for current pricing.

## How to participate

For the full step-by-step walkthrough and a realistic earnings picture, see [How to Earn with Grass](https://ownyourmind.ai/use/how-to-earn-with-grass/).

**Run a Grass node.** The most accessible entry point. Install the browser extension, desktop app, or Android app. The node passively shares unused bandwidth to scrape public web data. No staking, no technical knowledge. Earn Grass Points based on bandwidth contribution, uptime, and location. Points convert to GRASS tokens during airdrop distributions. Season 2 claims went live on 23 July 2026 through the newly launched Grass-native wallet, with those rewards paid in USDC. Requires a residential IP (data centre IPs are not eligible).

**Buy a Grasshopper.** For dedicated bandwidth sharing without tying up a computer or phone. Grasshopper is a palm-sized aluminium device that plugs into your router via Ethernet and runs 24/7 as a persistent node. FCC and ICES-003 certified, fully encrypted in transit, physically isolated from your home network. Handles only Grass traffic. No access to personal devices or data. Pricing not yet disclosed; announced with a waitlist in mid-2025, but shipping has been delayed by tariff-driven supply chain constraints and actual delivery status remains unclear.

**Stake GRASS.** Delegate tokens to router operators via grassfoundation.io/stake. Routers with more delegated stake receive more network traffic, generating higher rewards. Approximately 7.49% APR, per Stakin. No minimum stake. 7-day unstaking period.

**Operate a router.** Advanced participation. Routers relay bandwidth between nodes and validators. Earn rewards from the 30 million Router Rewards allocation plus commission on delegator rewards. Hardware requirements not publicly specified. Details limited.

**Enterprise access.** AI companies access Grass data through the Foundation. Approximately 20 current customers. Vetting process and service agreement required. Pricing not publicly disclosed.

## Honest assessment

### What works

The reported scale is large. Grass claims over 8.5 million users across 190 countries and last reported 8.3 million active nodes in September 2025, which would make it one of the largest DePIN networks by participant count. No independent figure exists to check it against: the app is login-gated, one centralised validator measures node counts, and the only on-chain population is the 2.8 million wallets eligible for Airdrop One. Grass said in September 2025 that its 3 petabytes per day of data retrieval is driven by customer demand.

Enterprise traction now has a number on it. In September 2026 Grass published a review by Regen Financial covering every work order at Grass DataCo from 14 March 2025 to 30 June 2026: $32.1M booked, of which $27.6M (86%) was traced to cash receipts. Per the review, the rest is mostly timing: $2.7M not yet invoiced on long-term contracts and $1.8M invoiced but unpaid.

The VALID dataset collaboration with LAION and Ontocord produced a 14-terabyte open dataset (CC-BY) that demonstrates the data pipeline converts bandwidth into usable AI assets. ClipTagger-12b is a published model with Apache-2.0 licensing. These are verifiable outputs, not promises.

The airdrop model deserves credit. 2.8 million recipients receiving 10% of supply with no lockup is one of the widest distributions in Solana history. The 500-point minimum threshold meant anyone who ran the extension for roughly 100 hours could qualify. This created a credible base of millions of token holders, even if many were airdrop farmers.

Barrier to entry is near zero. Install an extension, earn tokens. No staking minimum, no technical skills. That is how Grass reached the 8.5 million participants it reports. The Android app (launched July 2025) extended access, and Grasshopper (a dedicated hardware node announced mid-2025, though shipping has been delayed by tariff-driven supply chain issues) adds a plug-and-forget option for users who want 24/7 uptime without leaving a computer running.

AppEsteem cybersecurity certification with continuous monitoring by Avast, Microsoft Defender, and McAfee provides third-party validation that the software isn't malware, addressing the most obvious concern about bandwidth-sharing applications.

### What doesn't work

The code is entirely closed-source. Wynd-Network's GitHub has zero public repositories. You can't read the node code, the router code, the validator code, or the ZK processor code. Every technical claim, the zero-knowledge proofs, the data provenance, the separation of personal browsing from network traffic, is unverifiable. For a project describing itself as a decentralised data layer, this is the most fundamental credibility gap possible.

Governance is non-functional. The documentation references DAO governance and 22.8% of supply allocated to the Foundation growth bucket "managed by DAO governance." In practice, no proposals have been submitted, no votes have occurred, no forum exists. All decisions are made by Wynd Labs and the Grass Foundation. The 22.8% Foundation allocation is team-controlled spending dressed in DAO language.

Users have no data sovereignty. You contribute bandwidth; the Grass Foundation captures all value from the resulting data. You have no control over what is scraped through your connection, no visibility into who buys the data, no say in pricing, and no share of enterprise revenue beyond token rewards. Your residential IP address is used as an exit node for web scraping. If a client scrapes restricted or copyrighted content through your connection, the legal exposure sits with you, not the Foundation. Web scraping legality varies by jurisdiction and is actively contested under GDPR, CFAA, and the EU Digital Services Act.

Validation is centralised. The documentation describes transitioning from a single centralised validator to a decentralised validator committee, but the current state is unclear. If one entity validates all data provenance claims, those claims rest on trust, not cryptographic verification.

Revenue disclosure is thin. The one set of figures comes from a review Grass commissioned and published, and Regen Financial states it gives "no opinion, conclusion, or assurance of any kind". Regen describes itself as an outsourced Web3 back-office firm; the full findings report is available only on request, and no client is named. The cash-traced share is lower for the first half of 2026 (78%) than for 2025 (92%), which is mostly timing: that period ended eleven weeks before the review, and some contracts carry 60-day payment terms.

None of it touches the chain. Grass says customer revenue lands at Grass DataCo, and there is still no public view of how the Foundation uses it.

### The risk

Exit node liability is the defining risk for participants. Your IP address is being used to scrape the web on behalf of unknown third parties. Despite client vetting and activity monitoring, the legal landscape for web scraping is shifting rapidly. The EU Digital Services Act, GDPR enforcement, and US CFAA precedents are creating a more restrictive environment. If regulators decide that bandwidth contributors are facilitating unauthorised data collection, millions of users face potential exposure.

The insider token concentration creates sustained sell pressure. 47.2% to investors and team, with the October 2025 cliff having released approximately 181 million tokens. Monthly unlocks continue through 2028-2029. At close to 90% below ATH, the token has already absorbed significant selling, but hundreds of millions of tokens remain locked.

Single-team dependency is acute. Wynd Labs controls all development, infrastructure, and client relationships. One publicly doxxed team member. If the team disappears or pivots, the network has no fallback. There's no open-source code to fork, no community governance to redirect, and no decentralised infrastructure to persist independently.

The Sovereign Data Rollup has been "launching soon" for over a year. Without it, the ZK proof claims are marketing rather than verifiable technology. Shipping the rollup with open-source code would address the two largest credibility gaps simultaneously.

### My position

I hold GRASS from the airdrop distribution. The reported scale is large, though only Grass can measure it, and the first revenue figures make the enterprise traction more than plausible. I'm holding the airdrop position rather than adding to it: the code is entirely closed-source, governance does not exist in practice, users have no sovereignty over the data their bandwidth produces, and the revenue figures rest on one review with no audit opinion. I would add meaningfully if the core code is open-sourced, if governance materialises with binding proposals and votes, if the Sovereign Data Rollup ships with verifiable ZK proofs, and if revenue reporting becomes audited and recurring. The business works; the decentralisation does not.

## Freedom Score: 37/100

Grass scores 37/100 (F grade). Full methodology at [Freedom Score Methodology](https://ownyourmind.ai/resources/freedom-score-methodology/).

**Infrastructure decentralisation (10/20):** 8.3 million active nodes across 190 countries (Grass's last reported figure, September 2025) is distributed infrastructure at the edge. Nodes are permissionless; anyone with a residential IP can participate. However, validation is currently transitioning from a single centralised validator with the end state unclear. Routers are operated by approved infrastructure providers, not permissionlessly. Data storage is centralised on Backblaze B2 (2-3 petabytes daily). The Sovereign Data Rollup is planned but not confirmed live. Core scraping coordination, client management, and data delivery are controlled by Wynd Labs and the Grass Foundation. Permissionless at the edge, centralised at the core.

**Governance decentralisation (4/20):** DAO governance is referenced in documentation with 22.8% of supply allocated to the Foundation growth bucket "managed by DAO governance." In practice, no governance proposals have been submitted, no voting has occurred, and no governance forum exists. All decisions are made by Wynd Labs and the Grass Foundation. The 47.2% insider allocation (investors plus team) and Foundation-controlled 22.8% concentrate power. No evidence of meaningful community governance.

**Token distribution fairness (7/15):** Airdrop One (10% of supply to 2.8 million recipients with no lockup) was one of the widest distributions in crypto history. Future Incentives (17%) and Router Rewards (3%) provide ongoing community allocation totalling 30%. However, insider allocation is 47.2% (25.2% early investors plus 22% team). The 22.8% Foundation allocation is team-controlled. No public sale was held; funding came exclusively from VC rounds. The wide airdrop partially offsets the heavy insider concentration.

**Censorship resistance (6/15):** Nodes are permissionless and can contribute bandwidth from any residential connection. However, enterprise clients must go through a vetting process with service agreements; the Foundation controls all data access. Clients can be "immediately banned for activity deemed sketchy." Staking contracts include manual multisig-controlled slashing capability. Certain jurisdictions are excluded from token eligibility. The browser extension and apps can be removed from app stores. Permissionless contribution, controlled consumption.

**Data sovereignty (5/15):** Users contribute bandwidth but do not own or control the scraped data. The Grass Foundation captures all value from data and sells it to enterprise clients. Contributors have no control over what is scraped through their connection. Users act as exit nodes with their residential IP addresses used for third-party scraping. ZK proofs create data provenance records, but these are unverifiable with closed-source code. The Foundation has full visibility into all scraped data. Despite privacy safeguards and minimal personal data collection, the fundamental model is: users provide the resource, the Foundation captures the asset.

**Open source and transparency (5/15):** Wynd-Network GitHub has zero public repositories. GrassFoundation has one repository (GrassDownload, HTML). Node software, router software, validator software, and ZK processor are all proprietary. The Cantina security competition (June-July 2024, $15,000 prize pool, 116 findings) reviewed three Solana programmes but these are not publicly accessible. On the positive side, the VALID dataset is CC-BY, ClipTagger-12b is Apache-2.0, AppEsteem certification provides third-party security monitoring, and the token contract is on-chain and verifiable. Financial transparency is limited: the first revenue figures arrived in September 2026 through a commissioned review with no audit opinion, and Foundation spending remains undisclosed.

### Path to improvement

Three changes would materially increase Grass's score:

1. **Open-source the core code.** Zero public repositories is the single biggest credibility gap. Open-sourcing the node software, router software, and ZK processor would allow independent verification of every technical claim: the zero-knowledge proofs, the traffic separation, the data provenance. It would also enable community contributions and reduce single-team dependency. This is the highest-impact change Grass could make.
2. **Ship functional governance.** Launch a governance forum, publish the first proposals, run the first votes. The documentation already describes DAO governance; deliver it. Even modest governance, letting token holders vote on Foundation spending priorities or router selection criteria, would demonstrate the project is serious about decentralisation. Currently it is pure narrative.
3. **Make revenue and data transparency routine.** The September 2026 review was a first step. Audited figures on a fixed schedule, the number of enterprise clients, and how Foundation funds are allocated would complete it. If GRASS is "the primary vehicle for value accrual" as stated, token holders deserve visibility into what value is being accrued and where it flows.

## Returns Score: 55/100

GRASS scores 55/100 (C grade). Full methodology at [Returns Score Methodology](https://ownyourmind.ai/resources/returns-score-methodology/).

**Token utility (12/20):** GRASS is required for Live Context Retrieval (LCR), the enterprise product where AI models pull real-time web data to supplement prompts. Fiat payments from enterprise clients convert to GRASS, creating buy pressure proportional to usage. Users earn GRASS through bandwidth sharing. Staking to routers yields approximately 7.49% APR, per Stakin. The utility exists but is narrow: the token is a settlement layer for data access and a reward mechanism for bandwidth contribution. There's no governance utility in practice (despite documentation claiming it), and the staking mechanism lacks slashing enforcement, which reduces its economic credibility.

**Value accrual (8/20):** The thesis is that enterprise demand for AI training data drives GRASS purchases. Approximately 20 enterprise customers, including "the largest AI labs" and "seven-figure revenue clients," theoretically create demand-side pressure. Grass now publishes a revenue figure ($32.1M booked over 15.5 months, per its commissioned review), but no client names and no on-chain trail for enterprise payments. Token holders are still asked to trust that value accrues to the token, because the mechanism by which Foundation revenue translates to GRASS demand is entirely opaque. There is no burn mechanism, no fee distribution, and no verifiable buyback programme.

**Supply dynamics (8/20):** GRASS has a 1 billion total that its live mint authority could in principle exceed, but the distribution beneath it is heavily insider-weighted. Investors hold 25.2% and team 22%, for a combined 47.2% insider allocation. The Foundation controls an additional 22.8%. The October 2025 cliff released approximately 181 million investor tokens, and a further 55 million tokens unlocked to contributors on 28 February 2026. Monthly linear unlocks continue through 2028-2029. Hundreds of millions of tokens remain locked. The wide airdrop to 2.8 million wallets was notable, and Season 2 claims went live on 23 July 2026 via the newly launched native wallet, paid in USDC rather than new GRASS emissions. Combined community allocation (30%) doesn't offset 47.2% insider concentration regardless. At close to 90% below ATH, much of the selling may be absorbed, but the overhang from ongoing unlocks creates sustained downward pressure.

**Revenue sustainability (15/25):** This is the strongest element of Grass's returns case. AI companies paying for data derived from 3 petabytes of daily web scraping is a legitimate business with growing demand as foundation models scale. Grass-commissioned figures put booked revenue at $17.6M for 14 March to 31 December 2025 and $14.5M for the first half of 2026, with $27.6M of the combined $32.1M traced to cash. They put numbers where there were none, but they're unaudited and we can't re-check them, so this score doesn't rise on them.

The VALID dataset (14 terabytes, CC-BY) and ClipTagger-12b (Apache-2.0) are verifiable outputs that demonstrate the data pipeline works. The enterprise traction appears more durable than most DePIN revenue claims. The concern is that sustainability depends entirely on Wynd Labs and the Grass Foundation maintaining enterprise relationships. If the team disappears, the revenue disappears, because there is no open-source code to fork and no decentralised infrastructure to persist independently.

**Liquidity and access (12/15):** GRASS is spot-listed on Bybit, Gate.io, KuCoin, Kraken, Bitget, Bithumb, OKX (spot listing 24 April 2026) and Coinbase (spot trading from 26 August 2026). Reported volume is concentrated: on 23 September 2026 Aivora Exchange, a little-known venue, carried 47.6% of 24h volume in CoinGecko's tickers, with LBank a further 21.0%.  Binance Futures perpetual contracts (up to 75x leverage) add derivatives liquidity, though a Binance spot listing has not materialised despite community expectations. The exchange coverage is strong, though the venue concentration means headline volume overstates depth. The 2.8 million airdrop recipients created one of the widest holder bases in DePIN. DEX liquidity on Solana supplements the centralised exchange coverage. For a project under two years past TGE, the exchange access is well ahead of most peers.

### Path to improvement

Three changes would materially increase Grass's returns score:

1. **Audit the revenue and publish Foundation financials.** Grass took the first step in September 2026 with a commissioned revenue review, but it carries no audit opinion and covers a single window. Audited quarterly revenue and Foundation spending reports would move the investment thesis from trust-based to evidence-based.
2. **Implement a verifiable token burn or buyback mechanism.** There is currently no mechanism linking enterprise revenue to token value. A burn on LCR payments, an on-chain buyback programme, or a fee distribution to stakers would create a direct, verifiable connection between network usage and token economics. Without this, GRASS is a reward token with no reflexive demand loop.
3. **Reduce insider unlock pressure.** The 47.2% insider allocation with ongoing linear unlocks through 2028-2029 creates years of sustained sell pressure. Extending vesting schedules, implementing performance-based unlock milestones, or converting a portion to staking-locked tokens would signal long-term alignment and reduce the overhang that weighs on price discovery.

## Score change log

| Date | Score | Change | Reason |
| - | - | - | - |
| 2026-10-09 | Data | N/A | Contents API moved from planned to launched (8 October, access by request). Scores unchanged. |
| 2026-10-07 | Data | N/A | Network figures re-read at source: "8.5 million monthly active nodes" restated as Grass's homepage claim of over 8.5 million users, with 8.3 million active nodes and 3 PB a day dated to its September 2025 update. Scores unchanged. |
| 2026-10-04 | Data | N/A | Added Grass's planned Contents API, Search API and multimodal grounding from its 28 September post, worded as intent. Scores unchanged. |
| 2026-10-04 | Data | N/A | Added Multicoin Capital's 29 September investment in Grass (amount undisclosed). Scores unchanged. |
| 2026-09-24 | Data | N/A | Body sweep: the "scale holds up" passage now reads 8.5M as Grass's unverifiable claim, Coinbase and the Aivora volume concentration added to liquidity, ATH drawdown refreshed to ~89%, staking APR attributed to Stakin. Scores unchanged. |
| 2026-09-24 | Data | N/A | The description and summary stated Grass's 8.5M figure as fact; it is Grass's own claim with no independent figure, and now reads as one. Scores unchanged. |
| 2026-09-23 | Data | N/A | Supply wording corrected: the 1B total is policy, since a Squads multisig still holds mint authority. Scores unchanged. |
| 2026-09-23 | Data | N/A | Integrity records re-cut: 2 claims now quote Grass's own pages verbatim and 5 are our own measurements. Scores unchanged. |
| 2026-09-23 | Data | N/A | Grass published a commissioned Regen Financial revenue review: $32.1M booked, $27.6M traced to cash (Mar 2025 to Jun 2026), no assurance given. Revenue-opacity prose updated; scores unchanged because the figure is unverifiable by us. |
| 2026-09-05 | Data | N/A | Coinbase opened full GRASS-USD spot trading on 26 Aug 2026, five days after adding GRASS to its listing roadmap. Added to the Liquidity & Access evidence. Kraken was already on that list and is US-regulated, so this widens US access rather than creating it; Liquidity & Access is a candidate move at the next review, held to the higher bar because we hold this one. |
| 2026-08-12 | Data | N/A | WS3 Solana on-chain reconciliation (1->4 verdicts). Supply matches ~1B but is not enforced: the mint authority is a live single-keypair wallet (established). No freeze authority (established). Open-source confirmed CLOSED (Wynd-Network 0 public repos; established). Scores unchanged. |
| 2026-07-25 | Data | N/A | The Grass-native wallet launched 23 July 2026 and Season 2 (Stage 2) reward claims went live the same day, correcting the three "wallet has not launched as of May 2026" body references. Season 2 rewards are paid in USDC, not new GRASS emissions (verified via Coingabbar, 23 July). No score change. |
| 2026-05-03 | Editorial | N/A | Season 2 wallet correction: article incorrectly presented native-wallet distribution as live; the wallet has not launched as of May 2026. OKX spot listing (24 April 2026) added to exchange coverage. No score change. |
| 2026-04-06 | Data | N/A | Updated Grasshopper shipping status: delayed by tariff-driven supply chain constraints, delivery timeline unclear. |
| 2026-03-24 | Data | N/A | Removed stale circulating supply figure and volume. Verified against CoinGecko. |
| 2025-03-03 | Both | N/A | Initial publish. Freedom 37/100, Returns 55/100. |