Convex Backend Pricing Explained: Function Calls, Storage & TCO Calculator

Reactive Database Architecture, WebSocket Invalidation & Serverless Billing Math

Convex re-engineers backend persistence around reactive, deterministic TypeScript functions. Unlike traditional relational or document databases where clients continuously poll REST endpoints or manually configure caching layers, Convex establishes persistent WebSockets between clients and its transactional database engine. Whenever database state changes, Convex uses dependency tracking to re-execute subscribed query functions automatically and stream minimal binary diffs to connected frontend clients with sub-15ms sync latencies.

Deterministic Query Function with Built-in Vector Search

Define end-to-end type-safe queries combining transactional records and vector search in TypeScript:

import { query } from "./_generated/server";
import { v } from "convex/values";

export const searchArticles = query({
  args: { queryEmbedding: v.array(v.float64()), tenantId: v.string() },
  handler: async (ctx, args) => {
    const results = await ctx.db
      .vectorSearch("articles", "by_embedding", {
        vector: args.queryEmbedding,
        limit: 10,
        filter: (q) => q.eq("tenantId", args.tenantId)
      });
    return results;
  }
});

Transactional Serializability & Optimistic Concurrency Control (OCC)

All Convex mutations execute under full ACID serializability via Optimistic Concurrency Control (OCC). If two concurrent client mutations modify the same document simultaneously, Convex automatically retries the conflicting mutation against the freshly committed state without application deadlocks. This guarantees zero race conditions in high-concurrency payment and inventory systems.