IPFS's content-addressing (files by CID hash, not URL) means data is retrievable as long as one node pins it — no 404s. For Web3, this solves NFT metadata permanence: centralized server failures broke millions of NFTs. Filecoin's incentivized storage + IPFS retrieval provide persistence + availability together. Pinata and Lighthouse simplify pinning without running nodes. For AI model weights, public archives, and on-chain asset references, IPFS is the standard decentralized storage layer.
IPFS's content-addressing (files by CID hash, not URL) means data is retrievable as long as one node pins it — no 404s. For Web3, this solves NFT metadata permanence: centralized server failures broke millions of NFTs. Filecoin's incentivized storage + IPFS retrieval provide persistence + availability together. Pinata and Lighthouse simplify pinning without running nodes. For AI model weights, public archives, and on-chain asset references, IPFS is the standard decentralized storage layer.
Web3 dApp Usage
IPFS/Filecoin Web3 adoption, 2026Filecoin Data Secured
Filecoin network, 2025Storage Market (2024)
Decentralized storage market researchMarket Growth (2034)
22.4% CAGR projection to 2034Over 60% of Web3 dApps use IPFS/Filecoin for metadata and large assets — the ecosystem standard for off-chain decentralized storage
Content addressing via CID (Content Identifier) means the same content always has the same address — data deduplication and integrity verification built into the protocol
Filecoin's incentivized persistence layer secures 2.1 exbibytes of data across 7.6 exbibytes raw capacity — storage providers are economically compensated for keeping data available
No single point of failure — files pinned across multiple nodes remain available even if individual nodes go offline; no server to hack, DDoS, or take down
Decentralized storage market: $622.9M (2024) → $4.5B (2034) at 22.4% CAGR — growing institutional demand for censorship-resistant data persistence
Pinata, Lighthouse, and Filebase provide developer-friendly IPFS APIs — upload files, get CIDs, and manage pinning without running IPFS nodes or understanding DHT internals
IPFS integrates natively with Ethereum, Solana, and every major blockchain — smart contracts reference IPFS CIDs for off-chain data permanence proofs
Filecoin Onchain Cloud (January 2026) launches compute and verifiable retrieval layers — positioning as a full-stack decentralized AWS alternative for AI training data and archival storage
IPFS is the right storage choice for NFT projects requiring metadata permanence beyond centralized server lifetime, Web3 applications storing large assets (artwork, documents, datasets) that smart contracts reference by hash, open-source projects needing censorship-resistant distribution, and organizations archiving public-good data permanently. For applications requiring fast retrieval SLAs and centralized governance, traditional cloud storage (S3, GCS) remains simpler — IPFS's strength is permanence and decentralization, not raw retrieval speed.

ERC-721 and ERC-1155 NFT contracts store IPFS CIDs on-chain — the artwork image and JSON metadata live on IPFS permanently, not on a server that can be shut down. Metaplex on Solana uses the same pattern.
dApps storing user-generated content (posts, profiles, uploaded files) use IPFS to ensure content is accessible even if the application server goes offline — users retain access to their data through any IPFS gateway.
Filecoin's verified storage deals provide cryptographic proof that AI training datasets are stored intact — suitable for organizations sharing datasets publicly that need immutable, auditable storage provenance.
Research papers, open-source code, government records, and cultural archives stored on IPFS/Filecoin remain accessible regardless of institutional changes — Protocol Labs' Filecoin Foundation archives important human knowledge.
Fully decentralized dApp frontends deploy HTML/CSS/JS bundles to IPFS — accessible via Cloudflare's IPFS gateway, ENS .eth domains, and IPNS for mutable updates while preserving content-addressed history.
Smart contracts that need to reference large files (certificates, audit reports, evidence files) store the IPFS CID on-chain — providing an immutable pointer to off-chain content without expensive on-chain storage.
We believe in honest communication. Here are scenarios where alternative solutions might be more appropriate:
Private, access-controlled documents requiring authentication before retrieval — IPFS is public by default; private data requires encryption before upload, not access control at the storage layer
Applications requiring sub-100ms retrieval SLAs — public IPFS gateway latency varies by content popularity and node proximity; AWS S3 provides consistent low-latency retrieval for performance-critical data
Frequently updated data — IPFS content is immutable by CID; mutable data requires IPNS or sequential CID references, adding complexity vs traditional databases
Regulated data with geographic storage requirements — IPFS distributes data across global nodes; organizations with EU data residency requirements cannot guarantee data stays within jurisdictional boundaries
We're here to help you find the right solution. Let's have an honest conversation about your specific needs and determine if IPFS is the right fit for your business.
10,000-piece generative art NFT collection stores all metadata JSON files and artwork PNG images on IPFS via Pinata — CIDs embedded in ERC-721 contract's tokenURI. When OpenSea or any NFT marketplace queries the token, it resolves the IPFS CID to retrieve immutable metadata that can never be changed by the collection creator.
Example: Pinata IPFS: 10K NFT JSON metadata + artwork, CIDs on-chain in ERC-721 tokenURI, OpenSea/Blur compatible
DeFi protocol deploys its governance dashboard as a static build to IPFS — accessible via an IPFS gateway even if the team's servers go down. ENS domain (uniswap.eth) resolves to the current IPFS CID via IPNS content hash record, providing a human-readable address for the decentralized frontend.
Example: IPFS + ENS: governance frontend deployed to IPFS, uniswap.eth → IPNS content hash, accessible via Cloudflare gateway
Research institution stores 50TB AI training dataset on Filecoin with verified storage deals — cryptographic proofs (Proof-of-Replication, Proof-of-Spacetime) verify data is stored intact across multiple storage providers. Retrieval via IPFS CID; on-chain deal ID provides auditable storage receipt.
Example: Filecoin verified deal: 50TB AI dataset, Proof-of-Replication + Proof-of-Spacetime, IPFS retrieval
Supply chain contract stores IPFS CIDs for inspection certificates and quality reports on-chain — smart contract events emit the CID when goods transfer custody. Verifiers query the blockchain for the CID, retrieve the document via IPFS, and verify the content hash matches what's stored on-chain.
Example: Smart contract IPFS CIDs: inspection certificates stored on IPFS, hash verified against on-chain event log
Social dApp stores post content on IPFS via Lighthouse — the smart contract records the CID and author address. Users retrieve post content from IPFS; the blockchain provides authorship proof and ordering. Moderation is community-governed on-chain, not controlled by a central platform.
Example: Lighthouse IPFS: post content stored off-chain, CID + author on-chain, community moderation via smart contract
AI startup stores training data on Filecoin's verifiable storage and uses Filecoin Onchain Cloud (launched January 2026) to access compute for model training — all data provenance and compute jobs recorded on-chain for auditability. Eliminates dependence on centralized cloud providers for AI training workflows.
Example: Filecoin Onchain Cloud: AI training data + compute jobs, on-chain provenance, no centralized cloud dependency
Every technology has its strengths and limitations. Here's an honest assessment to help you make an informed decision.
IPFS CIDs are cryptographic hashes of content — the same file always has the same CID, and the CID verifies integrity. Data can never be silently modified; only new content generates new CIDs.
Files pinned across multiple IPFS nodes remain accessible even when individual nodes go offline. Unlike HTTP URLs pointing to one server, IPFS CIDs resolve to any node that has the content — geographic redundancy built into the protocol.
Filecoin's incentivized storage market pays storage providers to keep data available. Unlike IPFS alone (which relies on voluntary pinning), Filecoin storage deals create economic guarantees for data persistence over specified durations.
Smart contracts store IPFS CIDs as string references — linking immutable on-chain ownership records to immutable off-chain content. NFTs, DAOs, and DeFi protocols use this pattern to reference large data without expensive on-chain storage.
Pinata, Lighthouse, and Filebase provide HTTP APIs for uploading content and receiving CIDs — developers never run IPFS nodes, manage DHT, or understand distributed systems internals. Integration is as simple as uploading to S3.
IPFS is an open protocol with implementations in JavaScript (@helia/ipfs), Go (Kubo), and Rust — no vendor lock-in. Content pinned with Pinata can be retrieved via any IPFS gateway, including Cloudflare's public gateway.
IPFS retrieval speed depends on content popularity and node proximity. Popular content cached in Cloudflare's IPFS gateway resolves in milliseconds; unpopular content on few nodes can take 5-30 seconds for first retrieval.
We configure dedicated IPFS gateway infrastructure (Pinata's private gateway, Cloudflare R2 + IPFS) for production applications requiring consistent retrieval performance. For latency-sensitive content, we serve from centralized CDN while maintaining IPFS as the canonical, verifiable source.
Content only persists while at least one node pins it. Nodes may unpin content to free storage — without Filecoin storage deals or a paid pinning service, IPFS content can become unretrievable when all pins are removed.
We always pair IPFS with a paid pinning service (Pinata, Lighthouse) or Filecoin storage deal for production NFT and dApp content. We also implement multiple pinning providers for redundancy — if Pinata removes a pin, Lighthouse's copy remains accessible.
IPFS content is public by default — anyone knowing the CID can retrieve the content. Private data must be encrypted before upload, adding key management complexity.
We encrypt sensitive content using client-side encryption (libsodium, AES-256) before IPFS upload — the CID references encrypted ciphertext, not plaintext. Decryption keys are stored separately (user wallet, secure keystore) and never on IPFS.
IPFS content is immutable — updating content requires uploading a new file and getting a new CID. Applications needing mutable state (user profiles, document versions) must manage CID references at the application layer.
We use IPNS (InterPlanetary Name System) for mutable IPFS pointers where the application needs updateable references. For high-update-frequency data, we store IPFS CIDs in a database or smart contract rather than IPNS, providing application-level mutability with IPFS immutability for each version.
Every technology has its place. Here's how IPFS compares to other popular options to help you make the right choice.
Use Ethereum + IPFS together — they're complementary layers, not alternatives. Ethereum provides the ownership, identity, and transaction layer; IPFS provides the off-chain content storage layer. The standard NFT pattern is: mint ERC-721 on Ethereum with tokenURI pointing to IPFS CID.
Use Web3 frontend libraries (ethers.js/viem/wagmi) for blockchain interaction AND IPFS libraries/APIs (Pinata, Helia) for content storage. A complete NFT mint involves both: upload metadata to IPFS via Pinata API, receive CID, then pass CID to smart contract mint function via ethers.js.
For Solana NFT metadata, Arweave is often preferred over IPFS for its pay-once permanent storage (no ongoing pinning required). For cross-chain applications or when IPFS gateways are already integrated, IPFS works identically with Solana/Metaplex via standard URI references. Use IPFS + Pinata for flexible Solana storage; Arweave for permanent-by-default NFT collections.
We've built IPFS integrations for NFT collections storing 100,000+ assets, decentralized frontends serving thousands of daily users, and enterprise document archiving systems with on-chain IPFS CID references. We know which pinning service to use for each use case, how to design IPFS retrieval into production frontends without 30-second load times, and when Arweave's pay-once model is actually better than IPFS for your specific permanence requirements. We integrate IPFS properly — with redundant pinning, gateway fallbacks, and encryption for private content.
We upload NFT metadata JSON and artwork to IPFS via Pinata or Lighthouse, generating correctly formatted CIDs for ERC-721/ERC-1155 tokenURI and Metaplex standards — with redundant pinning across multiple services.
We create Filecoin verified storage deals for datasets, AI training data, and archival content requiring cryptographic proof of persistence — Proof-of-Replication and Proof-of-Spacetime verification included.
We configure IPFS gateway infrastructure for production dApps — dedicated Pinata gateways, Cloudflare IPFS gateway fallbacks, and CDN-cached content delivery ensuring consistent retrieval performance.
We deploy dApp static builds to IPFS with IPNS for mutable references, ENS content hash configuration for .eth domain resolution, and Cloudflare/Pinata gateway serving for fast global content delivery.
We integrate IPFS CID storage and retrieval into Solidity and Solana smart contracts — on-chain CID references for NFT metadata, document provenance, and off-chain data anchoring with content verification.
We implement client-side encryption before IPFS upload for sensitive content — AES-256 encryption, key management via user wallets or secure keystores, and IPFS storage of encrypted ciphertext only.
Have questions? We've got answers. Here are the most common questions we receive about IPFS.
IPFS (InterPlanetary File System) is a peer-to-peer protocol for storing and retrieving files. Unlike HTTP (which locates files by server address: http://server.com/file.jpg), IPFS locates files by content hash: ipfs://QmXyz... A CID (Content Identifier) is a cryptographic hash of the file content — the same file always has the same CID regardless of where it's stored. Requesting a CID from IPFS retrieves the content from any node that has it. Content addressing provides: integrity verification (CID proves content is unchanged), deduplication (same file never stored twice), and location independence (content accessible from any IPFS node).
IPFS is the protocol — it defines how files are content-addressed and transferred between peers. IPFS alone doesn't guarantee persistence; nodes only store what they choose to pin. Filecoin is an incentive layer on top of IPFS — it creates a marketplace where clients pay storage providers to keep data available, with cryptographic proofs (Proof-of-Replication, Proof-of-Spacetime) verifying data is stored correctly. For permanent storage: use Filecoin storage deals. For flexible storage with managed pinning: use IPFS with Pinata or Lighthouse pinning services.
Use at least two pinning services for redundancy. Pinata is the most popular — reliable API, generous free tier (1GB), fast gateways, and Pinata's dedicated private gateway provides consistent retrieval performance. Lighthouse provides permanent storage via Filecoin deals and supports token-gated encryption. Filebase is S3-compatible API backed by IPFS, ideal for teams already using AWS S3 tooling. For large collections (10K+ NFTs), we configure dual pinning on Pinata + Lighthouse with Filecoin storage deals for the permanent archive.
IPNS (InterPlanetary Name System) creates mutable pointers to IPFS content — an IPNS name resolves to a CID that can be updated over time. Use IPNS when your application needs a stable reference that points to the latest version of content: dApp frontends (the IPNS name stays the same; the underlying IPFS CID changes with each deployment), user profiles (one IPNS name per user, updated when profile changes), and software distribution (users always download the latest version from a stable IPNS address). Don't use IPNS for NFT tokenURIs — NFT metadata should be immutable.
Slow IPFS retrieval comes from content that isn't cached by popular gateways. Solutions: (1) Use a dedicated gateway (Pinata private gateway, Cloudflare R2 + IPFS) instead of public gateways. (2) Pre-warm content by fetching CIDs through Cloudflare's gateway after upload, getting content cached at edge. (3) Configure multiple gateway URLs with fallback — if Cloudflare is slow, try Pinata gateway. (4) For NFT art, cache IPFS images in your own CDN using the CID as the cache key — serve fast from CDN, verify against IPFS CID for integrity.
Standard ERC-721 metadata JSON: {name: 'Token #1', description: '...', image: 'ipfs://QmImageCID', attributes: [{trait_type: 'Color', value: 'Blue'}, ...]}. The image field should be an IPFS URI (ipfs://QmCID), not an HTTPS URL. Store images first, get their CIDs, then build metadata JSON with image CIDs, then upload metadata JSON and get its CID, then pass metadata CID to smart contract mint function. Never use https:// for NFT artwork — central server NFT artwork breaks when servers go down.
IPFS: content-addressed protocol, persistence requires active pinning, multiple provider options (Pinata, Lighthouse, Filecoin), flexible pricing model. Arweave: pay-once for permanent storage (200+ year endowment model), bundlr/Turbo for easy uploads, popular for Solana NFT metadata and permanent archival. Key difference: Arweave charges upfront for permanence with no ongoing fees; IPFS requires continuous pinning payments. Choose Arweave for set-and-forget NFT metadata; choose IPFS for flexible storage with multiple provider options and IPFS protocol compatibility.
IPFS itself is public — any node that knows a CID can retrieve the content. For private content: encrypt before uploading. Tools: (1) Lighthouse's access control encryption — encrypt with user's wallet signature, decrypt only with authorized wallet. (2) lit-js-sdk (Lit Protocol) — threshold encryption with on-chain access conditions (must hold this NFT to decrypt). (3) Client-side AES-256 encryption before upload — store decryption keys separately from IPFS. The CID references encrypted ciphertext; unauthorized parties get useless bytes.
Filecoin Onchain Cloud (launched January 2026) extends Filecoin beyond pure storage to a full-stack decentralized computing platform — adding compute layers, automated payment channels, and verifiable retrieval proofs. Traditional Filecoin: pay storage providers to hold data, prove via Proof-of-Spacetime. Onchain Cloud: combines storage with computation, allowing AI training jobs or data processing to run on Filecoin infrastructure with on-chain job receipts. Positioned as a decentralized alternative to AWS S3 + EC2 combination for privacy-preserving and censorship-resistant workloads.
We offer support covering: pinning service account management and redundancy monitoring, CID migration if a pinning service changes pricing or shuts down, Filecoin deal renewal before expiry, gateway configuration updates, IPFS library updates as @helia/ipfs and Pinata SDKs evolve, and performance monitoring for IPFS retrieval latency. We also provide content audit services — verifying all production CIDs are still pinned and accessible, before issues affect users.
Still have questions?
Contact UsExplore related technologies that work seamlessly together to build powerful solutions.

Most IPFS integrations we've seen treat pinning as an afterthought — upload to IPFS once, hope it stays available. We've seen popular NFT collections become unretrievable when the creator's single Pinata account expired. We design IPFS integrations with redundancy as a baseline: primary + backup pinning services, Filecoin storage deals for permanent content, and gateway configurations that don't serve 504s when a single provider has an outage. Your users' NFT artwork and application data will be available five years from now — that's the minimum bar we hold ourselves to.