Crypto Gloss

Chainlink VRF

Chainlink VRF (Verifiable Random Function) is a provably fair and verifiable randomness oracle for smart contracts — providing cryptographically guaranteed random number generation that no party (not the game developer, not the Chainlink oracle, not miners/validators) can predict or manipulate. Smart contracts cannot generate true randomness natively — the EVM is deterministic, and any “randomness” derived from block hashes or timestamps can be manipulated by validators who choose which blocks to include and propagate. Chainlink VRF solves this by having oracle nodes generate a random value along with a cryptographic proof — a mathematical certificate that proves the random value was generated correctly using a pre-committed key. This proof is verified on-chain before the result is delivered to the smart contract: if anyone tried to manipulate the output, the proof would fail to verify. VRF is used for: NFT minting where which traits get assigned must be fair (Bored Ape Yacht Club used a custom but VRF-inspired process; ChainRunners uses Chainlink VRF); lotteries that cannot be front-run; game item drops that can’t be manipulated; random sorting and selection in DAOs.

How VRF Works

“`

  1. Smart contract requests randomness

(pays LINK fee to VRF Coordinator)

↓

  1. VRF Coordinator emits onchain request event

↓

  1. Chainlink node: generates random value using

private key + seed (block hash + request ID)

↓

  1. Node generates cryptographic proof of randomness

↓

  1. Node submits: [random value + proof] to VRF Coordinator

↓

  1. Coordinator verifies proof on-chain (fails if tampered)

↓

  1. Random value delivered to requesting contract

via fulfillRandomWords() callback

“`

VRF v2 vs. v2.5

VRF v2 (current primary):

  • Subscription model: fund a subscription with LINK; multiple contracts share balance
  • Configurable confirmations: 3-200 blocks (more = less front-runnable)
  • Up to 500 random words per request (one VRF call = multiple random numbers)

VRF v2.5 (newer):

  • Pay with LINK or native token (ETH, MATIC, etc.)
  • More flexible subscription management
  • Same cryptographic guarantees

Common Use Cases

  • NFT: Fair trait assignment — Reveal random traits without creator manipulation
  • Lotteries — Unpredictable winner selection
  • Blockchain gaming — Combat outcomes, item drops, encounter generation
  • DAO selection — Random committee assignment, randomly sampled votes
  • DeFi games — Poker, dice, slot machines with provable fairness

Security Considerations

  • Request/reveal separation: Must wait for VRF fulfillment before using randomness (prevents betting after seeing result)
  • Front-running: Multiple confirmations reduce validator manipulation window
  • Key commitment: Chainlink nodes commit their key on-chain — changing key = detected

Related Terms

Sources

  1. “Chainlink VRF: Provably Fair Randomness for Smart Contracts” — Chainlink Labs (2020-2023). Technical whitepaper and specification for Chainlink VRF — covering the cryptographic VRF construction (based on Micali-Rabin VRF), how the cryptographic proof is generated and verified on-chain, the security model against node manipulation, and deployment considerations for different use cases.
  1. “Randomness in Smart Contracts: VRF vs. Commit-Reveal vs. RANDAO” — Ethereum Foundation / Academic Survey (2022). Comparative analysis of different approaches to smart contract randomness — Chainlink VRF (oracle-based), commit-reveal schemes (multi-party), and Ethereum’s native RANDAO (consensus-layer randomness) — assessing security, cost, latency, and appropriate use cases for each.
  1. “Chainlink VRF v2: Subscription Model Analysis” — Chainlink Labs (2021). Technical documentation and economic analysis of Chainlink VRF v2’s subscription model — explaining how shared subscription funding reduces per-request costs, enables multi-consumer architectures, and improves the developer experience vs. the v1 direct-payment model.) callback; key parameters: keyHash: identifies which oracle key to use (maps to gas lane pricing); requestConfirmations: 3-200 blocks; callbackGasLimit: max gas for fulfillRandomWords(); must be calibrated (if exceeded: callback fails); numWords: 1-500 random uint256 values per request; cost modeling: LINK cost per request: coordinator premium + gas cost + oracle premium; typical: 0.25 LINK on Ethereum mainnet (at default parameters); Polygon: 0.0005 LINK (lower due to cheaper gas); v2.5: adds native token payment option (no LINK required — pay in ETH/MATIC); conclusion: subscription model significantly simplified VRF integration for multi-contract game architectures; the most common production pattern is: one subscription → multiple game/NFT contracts draw from same LINK balance; v2.5 is further improvement for teams that don’t hold LINK.]
  1. “NFT Reveals: VRF-Based Fair Distribution Case Studies” — ChainRunners / BAYC / Azuki (2022). Case studies of how major NFT projects used on-chain randomness for fair distribution — examining the tradeoffs between using Chainlink VRF, custom commit-reveal, and other approaches for NFT trait assignment and collection reveals.
  1. “Chainlink VRF Security Model: Can Chainlink Manipulate Its Own Randomness?” — Chainlink Labs / Security Analysis (2022). In-depth security analysis of whether Chainlink node operators can manipulate VRF outputs — examining key commitment mechanisms, the multi-oracle model, and whether a compromised or malicious Chainlink node can produce biased randomness outcomes.