The End of the Public Chain 'Money-Splashing' Era

marsbitDipublikasikan tanggal 2026-04-13Terakhir diperbarui pada 2026-04-13

Abstrak

The article argues that the traditional L1 blockchain grant model for ecosystem funding is fundamentally broken, not just in need of minor adjustments. It claims that over the past four years, billions of dollars have been poured into grants from major L1s like NEAR, Avalanche, and Aptos, yet this has failed to produce lasting, breakthrough projects or sustainable ecosystems. Instead, it has created a cycle where teams optimize for grant applications and short-term KPIs rather than building viable products for real users. The model is criticized for fostering a "hostage" situation: foundations are trapped by their own metrics and feel compelled to disburse funds to show growth, while teams become dependent on grant cycles, shaping their roadmaps around fundable proposals rather than market needs. This leads to a proliferation of "zombie" projects that survive on grants but lack organic growth. The author contrasts this with direct equity and token investments, like those made by Solana Ventures, which create long-term, aligned partnerships. The piece concludes that the era of grants for cold-starting ecosystems is over. The future winners will be L1s that compete by attracting truly great companies through strategic investments and by being the best place to build, not the most generous grant committee.

Author: Dara_VC

Compiled by: Jiahuan, ChainCatcher

The L1 Grant model has failed. Not the kind of "needs fine-tuning" failure, but a complete failure in structure, philosophy, and incentive alignment.

And the L1 teams running these programs are either "unable to see the forest for the trees" because they are too close to it, or too afraid of the negative impact of stopping to admit it.

Where Did the Money Actually Go?

NEAR once announced an $800 million ecosystem fund, with $250 million specifically earmarked for ecosystem grants over the next four years. Prior to this, NEAR had already distributed over $45 million in grants to more than 800 projects.

Avalanche committed over $250 million in grants, dedicated to driving its ecosystem development.

Aptos runs milestone-based ecosystem grants ranging from $5,000 to $50,000, with payment-focused grants going up to $150,000. BNB Chain provides up to $200,000 per project.

Overall, across all major L1s, the capital funneled into this "grant machine" over the past four years amounts to hundreds of millions of dollars, possibly even exceeding a billion. As of 2024, there are over 50 active Web3 grant programs globally supporting various projects, including public goods, DeFi, tools, AI, and infrastructure.

You might think, with such a massive influx of capital, we should see a plethora of breakthrough companies, unicorn protocols, and ecosystems that genuinely retain liquidity and users long-term.

But that's not the case. TVL (Total Value Locked) churns and drains away, developers chase the next incentive mechanism. The impressive data reported to the board last quarter looks embarrassing six months later. And the question no one wants to ask out loud is: Where did all that money actually go?

In Defense of the Original Intent of the Grant Model

To be fair, the grant model once made sense. In 2020 and 2021, when L1s were genuinely trying to cold-start ecosystems from scratch, grants were a reasonable spark. You need developers before you have users, protocols before you have liquidity. Grants could ignite the initial flywheel.

In the early stages of Web3 development, ecosystem grants played a crucial funding role. They supported open-source contributions, incentivized participation in new protocols, and allowed teams to build MVPs (Minimum Viable Products) without immediate monetization pressure. Grants were ideal for ideation and experimentation.

"Spark" is the key word here. No one designed grants to be a permanent fuel source. But that's precisely what they have become in many ecosystems—a long-term "drip feed" that keeps projects on life support, never forcing them to learn to breathe on their own.

The proliferation of grant-dependent projects exposes key limitations. Grants often encourage short-term thinking, with teams optimizing for funding rounds rather than sustainable operation. Projects can get stuck in a death spiral of writing proposals and chasing grants, with less focus on building a viable user base or revenue-generating product.

The Hamster Wheel Trap of Spinning in Place

A team targets a mid-tier L1—one with a well-funded foundation, an active grants committee, and, crucially, lower competition for grant funds. They develop a product that fits the current grant wishlist: DeFi tools, a DEX, an NFT marketplace, some kind of "AI integration" (whatever that means this cycle).

They submit a slick proposal, hit the KPIs outlined in the milestone structure, receive funds in tranches, and generate activity metrics the foundation team can screenshot for their quarterly reports.

A small group of established teams repeatedly win the money, the opportunities, and the attention. Even in systems like quadratic funding, these same teams often dominate, crowding out newcomers.

Over time, the smarter teams figure out this cartel dynamic and simply play within it. They build relationships with the grants committee, become "known faces" in the ecosystem Discord, and position themselves as reliable recipients of funds, cycle after cycle.

Then, when the ecosystem hits its ceiling, when TVL stops growing, when real liquidity remains on Solana and Ethereum (because that's where the real users are), these teams do the rational thing. They start scoping out the next active ecosystem, port their code, write new proposals, and move on.

Grant programs measure success by the number of grants given or funds distributed, but that doesn't tell the whole story. The TVL charts tell the truth, the developer retention data tells the truth, the dead Discord channels tell the truth.

The Unspoken "Hostage" Dilemma

The grant model creates a strange dynamic that is rarely discussed candidly: it creates a hostage situation, and both sides are hostages.

The foundation becomes a hostage to its own metrics. They promised deployed capital, must report ecosystem growth to the board, and the easiest way to show growth is to offer more grants, more projects, prettier numbers.

The Ethereum Foundation funded 105 projects before realizing it needed to pause open applications. The volume itself became the problem, overwhelming a lean team and making it impossible to assess real long-term impact.

Even the most mature, credible ecosystem in the space, Ethereum, eventually had to stop and reflect... are we actually creating value, or just manufacturing activity?

The grantee teams are the other hostage. Once you're in the grant cycle, your organization structures itself around it. Your roadmap becomes grant proposals, your KPIs become whatever the committee wants to see. You stop making product decisions based on user needs and start making them based on what gets funded.

Web3 founders and developers must recognize that success is measured not just by funding rounds or community hype; long-term impact comes from building infrastructure and applications that stand the test of time. Grants can be the spark, but they must not be the fuel.

The tragedy is that genuinely talented teams get trapped. They could be creating real value, but instead, they're optimizing grant applications and networking in Telegram groups.

The Real Role of Direct Equity Investment

Let's contrast this with L1 venture arms writing real checks—investing in companies they genuinely believe in, with equity plus tokens.

Solana Ventures, the strategic investment arm of Solana Labs, has a clear mission to accelerate the development of the Solana blockchain itself, using its capital as leverage for ecosystem growth. The firm often co-designs development partnerships with game studios.

It's not just an investor; it's an infrastructure and go-to-market partner, helping teams build Solana-native game economies and integrations.

This is a fundamentally different relationship than a grant. When you take equity and tokens, you're betting that this company will matter. It changes everything about how you interact with them. You're on the same side of the table now.

You want them to find product-market fit (PMF), you want them to raise a real Series A, you want them to reach a billion-dollar valuation, because that billion-dollar valuation does more for your ecosystem's credibility, your token price, and the long-term narrative than 50 grant recipients combined.

A16z Crypto invested $50 million in the core Solana protocol Jito, trading capital for tokens with the explicit goal of fostering long-term alignment between the two companies. This is the way. Not a $50,000 grant with a milestone report due in 90 days.

This is betting on a company that will actually move the needle, with real skin in the game.

In 2025, global blockchain VC funding reached $35 billion, with firms like a16z Crypto and Pantera Capital leading multiple significant rounds. This is the pool L1 venture arms need to be competing in.

Developers Are Loyal to Users and Liquidity, Period

Another strategic mistake the grant model makes is this: it assumes developer loyalty can be purchased with non-dilutive capital. It cannot.

L1 activity in 2025 is bifurcating into roles: Solana, BNB Chain, and Hyperliquid capture massive speculative flows, while Ethereum consolidates its position as a settlement and data availability layer. The base layer continues to subdivide into specialized chains covering privacy, performance, and app-chain coordination, making interoperability and cross-chain routing increasingly important.

The best builders have figured this out. They're not loyal to a chain; they're loyal to users and liquidity. They go where the users are, where the exit potential is, where there's real volume and money movement.

And right now, it's a multi-chain reality. The winners of the next cycle will be protocols and apps with meaningful integrations across multiple chains.

Layer-1 chains raised approximately $2.71 billion between 2023 and 2025, with nearly 48% of the funding flowing to early-stage projects. Investors still back new execution environments but increasingly expect faster ecosystem delivery.

The market is getting smarter, and it's punishing ecosystems that rely on grants to prop up fake activity. Real throughput, real users, and real revenue are being rewarded now.

So what should L1 venture arms do? Get into the best companies early with equity, with tokens, with real strategic participation in a multi-chain portfolio, making integration with your chain part of their roadmap, not an optional extra.

If you back a company that is going to matter, you work with them to make your chain the natural home for their activity. You earn that loyalty by being the best technical environment to build their product, not by renting it with a grant check.

One Billion-Dollar Company vs. 200 Zombie Projects

Scenario A: You deploy $10 million in grants to 200 projects over two years. Your quarterly board report has DAU numbers, some GitHub activity, and a bunch of quotes from teams active on Discord optimizing for grant KPIs.

Two years later, half of those projects are either dead or have migrated to where the real liquidity is. Your TVL is stagnant, your developer retention data is awful. You report "funded 200+ projects" and pray no one asks for their whereabouts.

Scenario B: You take the same $10 million and deploy it into 10 truly promising companies in your ecosystem via direct equity plus tokens, with an integration roadmap that ties their success to your chain.

You provide real strategic support—not milestone nagging, but intros to hires, tokenomics design, go-to-market strategy. Three years later, one company is worth $1 billion, two others are worth $200 million each.

That billion-dollar company is the proof point that changes everything. It changes what other builders think about building in your ecosystem, what VCs think about writing checks in your ecosystem, what exchanges think about projects on your chain, what LPs think about your token.

The narrative gravity of one真正的 breakthrough project is immense, and its compounding effects are something 200 grant-surviving "zombie" projects could never achieve.

In Q1 2025 alone, blockchain and crypto startups raised $4.8 billion, the strongest quarter since late 2022. Startups that can demonstrate utility, compliance, and scalability attract not only capital but also strategic partners and long-term support.

Smart capital is already flowing to real companies that create real outcomes. L1 venture arms need to get into that flow, not run parallel grant programs that isolate them from it.

Kill the vanity metrics. Stop reporting the number of projects funded, start reporting portfolio company valuations, the TVL of your invested companies, and developer retention that's tied to organic growth, not incentive programs.

Draw a hard line between infrastructure grants and company investments. Some things are worth granting—genuine open-source public goods, core infrastructure, security research. These are true public goods that benefit everyone in the ecosystem and don't require a business model. But a DeFi protocol or a gaming app? That's a company. Invest in it like one.

Those Still Splashing Cash Will Be Washed Out

The L1 landscape in 2026 is vastly different from 2021. The total market cap of the L1 sector stabilized above $2.96 trillion, and competition has shifted from theoretical to practical applications, stablecoin payments, gaming, perp DEXs, creator tools, and app-specific chains. Winners are pulling away on throughput, fees, decentralization, and developer appeal.

The grant era made sense in the cold-start era. That era is over. What's left is the real competition for the best builders, the best protocols, the most real economic activity. You don't win that competition by being the most generous grants committee.

You win because you have a cohort of truly great companies that choose your chain (often among many options) because it's the best place to build, and they stay because you are the right long-term partner.

The L1s that figure this out in the next 18 months will look like visionaries. And those still running grant programs as their primary ecosystem development strategy will be exposed for what they are: mistaking activity for value creation and paying hundreds of millions of dollars for the self-deception.

Pertanyaan Terkait

QWhy does the author argue that the L1 grant model has completely failed?

AThe author argues that the L1 grant model has failed structurally, philosophically, and in terms of incentive alignment. It encourages short-term thinking, creates a cycle where teams optimize for funding rounds rather than sustainable operations, and results in projects that rely on perpetual 'life support' instead of achieving organic growth and product-market fit.

QWhat negative dynamic does the grant model create between foundations and recipient teams?

AThe grant model creates a 'hostage' dynamic where foundations become hostages to their own metrics, pressured to show growth by disbursing more grants and reporting positive data. Recipient teams become hostages by structuring their organizations around grant cycles, prioritizing fundable proposals over genuine user needs and sustainable product decisions.

QHow does direct equity investment differ from the grant model in fostering ecosystem growth?

ADirect equity investment, involving both equity and tokens, aligns long-term interests between the L1 and the company. It fosters a partnership where the investor actively supports the company's success, including product-market fit, future funding rounds, and integration with the L1. This creates more substantial, lasting value than short-term grant milestones.

QAccording to the author, what should L1 venture arms focus on instead of running grant programs?

AL1 venture arms should focus on making strategic equity and token investments in promising companies, integrating them into a multi-chain portfolio, and ensuring their chain becomes a core part of the company's roadmap. They should provide real strategic support, such as introductions for hiring, tokenomics design, and go-to-market strategies, rather than managing grant milestones.

QWhat is the key difference between 'vanity metrics' and genuine success indicators for an L1 ecosystem?

AVanity metrics, such as the number of funded projects or GitHub activity, often reflect artificial activity driven by incentives rather than organic growth. Genuine success indicators include portfolio company valuations, TVL from invested companies, and developer retention rates tied to real growth, not incentive programs. These reflect true value creation and ecosystem health.

Bacaan Terkait

Ulasan Pedas | Michael Saylor Ungkap 'Pernyataan Pria Playboy'; Nenek 60 Tahun 'Kena' Pemuda Setelah Kontraknya Liquidas

**Ringkasan:** **1. "Pernyataan Playboy" Michael Saylor:** Pendiri MicroStrategy, Michael Saylor, membuat klarifikasi kontroversial di BTC Prague. Ia menyatakan bahwa pesannya "jangan pernah menjual Bitcoin" hanya ditujukan untuk investor perorangan, bukan perusahaannya. MicroStrategy sebagai perusahaan tetap dapat menjual BTC jika diperlukan. Pernyataan ini memicu kritik keras dari komunitas crypto yang menganggapnya sebagai permainan kata dan standar ganda, mengingat sebelumnya Saylor sering menggembar-gemborkan komitmen hold jangka panjang. **2. Nenek 60 Tahun Tipu Pria Muda untuk Modal Trading Leverage:** Seorang nenek berusia 60 tahun di Beijing menipu seorang pria muda lebih dari 200,000 RMB (sekitar 450 juta Rupiah) dengan modus pacaran online. Ia menyamar sebagai "anak angkat" perempuan muda yang bekerja di kementerian, padahal itu adalah akun dirinya sendiri. Uang hasil penipuan tersebut ia gunakan untuk trading cryptocurrency dengan leverage 10x. Sayangnya, posisinya mengalami likuidasi (rugi total) saat pasar turun. Nenek tersebut akhirnya dihukum 4 tahun penjara karena penipuan. **3. Trader yang Kehilangan $45 Juta Hingga Hanya Tersisa $17 Ribu:** Seorang pengguna Reddit membagikan kisah pilunya. Melalui investasi dalam meme coin, nilai portofolionya pernah mencapai puncak $45 juta. Namun, karena greed (keserakahan) dan tidak mengambil profit, nilainya merosot drastis hingga hanya tersisa sekitar $17,000. Postingannya menjadi ruang "pengakuan dosa" dan peringatan bagi trader lain tentang pentingnya manajemen risiko, diversifikasi, dan disiplin mengambil keuntungan. **Kesimpulan:** Minggu ini di dunia crypto diwarnai oleh permainan kata dari tokoh besar, penipuan yang nekat, dan kisah kehancuran finansial akibat ketamakan. Semua cerita ini mengingatkan akan pentingnya kehati-hatian, skeptisisme, dan pengelolaan risiko yang baik dalam berinvestasi.

Foresight News12m yang lalu

Ulasan Pedas | Michael Saylor Ungkap 'Pernyataan Pria Playboy'; Nenek 60 Tahun 'Kena' Pemuda Setelah Kontraknya Liquidas

Foresight News12m yang lalu

Gemetarlah Manusia, AI Masih Berlari Kencang

Menggetarkan manusia, AI terus melaju dengan kecepatan penuh. Dalam acara BAAI Conference 2026, terlihat jelas bahwa seluruh industri, mulai dari model, perangkat lunak dan keras hingga produk, berupaya keras agar AI dapat "berlari" dari dunia digital ke dunia fisik. Di satu sisi, Scaling Law terus menunjukkan hasilnya, mendorong perkembangan model bahasa besar dan model multimodal. Industri AI telah memasuki tahap mengejar model dunia (world model), meskipun rute teknologi dan masalah data masih belum terselesaikan dan memerlukan eksplorasi setidaknya 3-5 tahun ke depan. Di sisi lain, terobosan dalam Agent mempercepat penerapan AI dalam skenario dunia nyata. Dengan Agent mencapai tahap yang dapat digunakan, industri sedang mempromosikan penerapannya dalam skenario seperti perawatan kesehatan dan rapat. Untuk membuat Agent dari "dapat digunakan" menjadi "mudah digunakan", kolaborasi perangkat lunak dan keras menjadi kunci. Pameran di konferensi ini didominasi oleh produsen chip AI terkemuka dari dalam negeri. Presiden BAAI Institute, Wang Zhongyuan, menyatakan bahwa AI sedang berevolusi secara mandiri. Dengan pematangan AI Coding dan percepatan penerapan Agent, AI telah meningkat dari menulis kode hingga menyelesaikan iterasi dan pembaruan produk secara mandiri. Model dunia telah menjadi medan pertempuran kunci berikutnya untuk model besar. Saat ini, belum ada konsensus teknis yang lengkap dalam industri, dan berbagai rute teknologi berkembang dengan caranya masing-masing. BAAI Institute mengklasifikasikan model dunia menjadi empat kategori dan sedang mengeksplorasi jalur kelima: integrasi pusat bahasa dan pusat representasi visual. Mereka juga memperkenalkan model dasar dunia universal yang sedang dalam pengembangan, "Wujie · Physis-v0.1", yang direncanakan akan open source setelah pelatihan selesai. Dari segi produk, Agent telah menjadi produk kunci bagi AI untuk memasuki kehidupan masyarakat. Tahun ini, Agent menjadi lebih proaktif dan mampu menangani tugas yang lebih kompleks. BAAI Institute meluncurkan empat Agent vertikal, termasuk asisten diagnosis untuk pencitraan resonansi magnetik jantung dan Agent pendengar rapat. Namun, untuk bergerak dari "dapat digunakan" ke "mudah digunakan", Agent masih menghadapi banyak tantangan teknis yang perlu dioptimalkan, seperti memori, orkestrasi, dan penyempurnaan kerangka kerja rekayasa (Harness) untuk meningkatkan kemampuan pemahaman dan eksekusi.

marsbit23m yang lalu

Gemetarlah Manusia, AI Masih Berlari Kencang

marsbit23m yang lalu

Sisi Lain Kekayaan Triliunan Dolar Musk: 85% Tak Bisa Dijual

Elon Musk menjadi triliuner pertama di dunia setelah valuasi SpaceX mencapai $1,77 triliun. Namun, kekayaan bersihnya yang diperkirakan melampaui $1 triliun sangat tidak likuid. Meski menguasai 85,1% hak suara, kepemilikan ekonomi aktualnya mungkin jauh lebih rendah, dan sebagian besar sahamnya memiliki hak suara super dengan pembatasan penjualan ketat. Ia diperkirakan hanya dapat mencairkan sekitar 1,6%-2,4% dari kekayaannya per tahun. SpaceX IPO dengan valuasi $1,77 triliun, tetapi hanya 4,2% dari total saham yang beredar di pasar sekunder, membuat harga saham sangat rentan terhadap volatilitas. Sekitar 4.400 karyawan diperkirakan menjadi jutawan kertas karena rencana kepemilikan saham, namun kekayaan mereka dikunci selama periode penahanan dan menghadapi kewajiban pajak. Sementara itu, divisi xAI SpaceX mencatatkan kerugian $64 miliar pada 2025, meskipun memiliki kontrak sewa komputasi senilai $260,4 miliar per tahun dari Anthropic dan Google. Pengeluaran modal tahunan xAI sekitar $308 miliar, menimbulkan pertanyaan tentang keberlanjutan bisnis ini. Analisis membandingkan potensi kekayaan karyawan jika Anthropic dan OpenAI go public, menunjukkan nilai yang jauh lebih tinggi daripada SpaceX karena struktur kepemilikan yang lebih tersebar, tetapi kelangsungan valuasi tinggi mereka di pasar publik masih belum teruji. Intinya, status triliuner Musk dan kekayaan karyawan SpaceX sebagian besar masih di atas kertas, dengan ujian sesungguhnya akan datang setelah periode lock-up berakhir dan tekanan jual memasuki pasar yang sangat tidak likuid.

链捕手33m yang lalu

Sisi Lain Kekayaan Triliunan Dolar Musk: 85% Tak Bisa Dijual

链捕手33m yang lalu

Selera Risiko Derivatif Kripto Terjun Bebas Seiring Outflow ETF Hantam Bitcoin

TL;DR: Apresiasi risiko di pasar derivatif kripto turun tajam menyusul penurunan harga spot hampir 20% pekan lalu, sejalan dengan arus keluar berkelanjutan dari ETF Bitcoin. Analisis Deribit Insights menunjukkan Indeks Apresiasi Risiko Block Scholes anjlok di bawah 0,05, mengindikasikan permintaan atas aset berisiko telah runtuh. Posisi defensif ini bertepatan dengan periode arus keluar terpanjang dari ETF Bitcoin spot sejak diluncurkan, yang menjadi pengukur sentimen institusional utama. Meskipun harga Bitcoin terkonsolidasi di atas $60.000, data opsi menunjukkan pemain pasar masih berhati-hati. Kemiringan risiko (skew) BTC 25-delta membaik dari level -19% menjadi hampir -9%, namun nilai negatif ini menunjukkan perlindungan dari penurunan harga masih lebih dihargai daripada eksposur kenaikan. Di pasar Ethereum, tingkat pendanaan (funding rate) telah negatif sejak 5 Juni, menandakan bias bearish di pasar perpetual swap. Hal ini mencerminkan sentimen yang tetap rapuh, dengan harga ETH masih turun 66% dari rekor tertinggi Agustus 2025. Secara keseluruhan, laporan ini menyimpulkan pasar kripto masih dalam mode perbaikan pasca penjualan besar-besaran, dengan posisi trader profesional yang lebih berhati-hati daripada percaya diri. Pemulihan berkelanjutan dalam arus ETF Bitcoin dan normalisasi tingkat pendanaan ETH perlu dipantau untuk sinyal kembalinya apresiasi risiko.

bitcoinist2j yang lalu

Selera Risiko Derivatif Kripto Terjun Bebas Seiring Outflow ETF Hantam Bitcoin

bitcoinist2j yang lalu

Trading

Spot
Futures

Artikel Populer

Cara Membeli ERA

Selamat datang di HTX.com! Kami telah membuat pembelian Caldera (ERA) menjadi mudah dan nyaman. Ikuti panduan langkah demi langkah kami untuk memulai perjalanan kripto Anda.Langkah 1: Buat Akun HTX AndaGunakan alamat email atau nomor ponsel Anda untuk mendaftar akun gratis di HTX. Rasakan perjalanan pendaftaran yang mudah dan buka semua fitur.Dapatkan Akun SayaLangkah 2: Buka Beli Kripto, lalu Pilih Metode Pembayaran AndaKartu Kredit/Debit: Gunakan Visa atau Mastercard Anda untuk membeli Caldera (ERA) secara instan.Saldo: Gunakan dana dari saldo akun HTX Anda untuk melakukan trading dengan lancar.Pihak Ketiga: Kami telah menambahkan metode pembayaran populer seperti Google Pay dan Apple Pay untuk meningkatkan kenyamanan.P2P: Lakukan trading langsung dengan pengguna lain di HTX.Over-the-Counter (OTC): Kami menawarkan layanan yang dibuat khusus dan kurs yang kompetitif bagi para trader.Langkah 3: Simpan Caldera (ERA) AndaSetelah melakukan pembelian, simpan Caldera (ERA) di akun HTX Anda. Selain itu, Anda dapat mengirimkannya ke tempat lain melalui transfer blockchain atau menggunakannya untuk memperdagangkan mata uang kripto lainnya.Langkah 4: Lakukan trading Caldera (ERA)Lakukan trading Caldera (ERA) dengan mudah di pasar spot HTX. Cukup akses akun Anda, pilih pasangan perdagangan, jalankan trading, lalu pantau secara real-time. Kami menawarkan pengalaman yang ramah pengguna baik untuk pemula maupun trader berpengalaman.

792 Total TayanganDipublikasikan pada 2025.07.17Diperbarui pada 2026.06.02

Cara Membeli ERA

Diskusi

Selamat datang di Komunitas HTX. Di sini, Anda bisa terus mendapatkan informasi terbaru tentang perkembangan platform terkini dan mendapatkan akses ke wawasan pasar profesional. Pendapat pengguna mengenai harga ERA (ERA) disajikan di bawah ini.

活动图片