diff --git a/skills/monid/SKILL.md b/skills/monid/SKILL.md
index 111a190..f613fe3 100644
--- a/skills/monid/SKILL.md
+++ b/skills/monid/SKILL.md
@@ -135,6 +135,7 @@ Each command supports `--help` for full usage. Here's what's available:
 | `monid runs get` | Get run status and results (`-r <runId>`, `-w` to wait) |
 | `monid runs stop` | Stop an in-progress run (`-r <runId>`). Not all runs can be stopped |
 | `monid balance` | Show current workspace balance |
+| `monid spend` | Summarize workspace spend by provider and endpoint |
 | `monid setup` | Complete CLI setup after installation (no API key required) |
 | `monid keys add` | Add an API key (`-k <key> -l <label>`) |
 | `monid keys list` | Show configured keys |
diff --git a/src/api/types.ts b/src/api/types.ts
index 83cb8c3..f0fce5f 100644
--- a/src/api/types.ts
+++ b/src/api/types.ts
@@ -348,6 +348,10 @@ export interface RunListItem {
   providerResponse?: ProviderResponse;
   price: Price;
   cost?: Cost | null;
+  /** Result items returned by the run (drives PER_RESULT billing). */
+  resultCount?: number;
+  /** Units actually charged for the run. */
+  billedUnits?: number;
   createdAt: string;
   startedAt?: string;
   completedAt?: string;
@@ -363,6 +367,34 @@ export interface RunsListResponse {
   usage?: Usage;
 }
 
+// --- Spend ---
+
+/** Aggregated run count and spend for one provider or endpoint. */
+export interface SpendBucket {
+  provider: string;
+  providerName?: string;
+  /** Present on endpoint buckets only. */
+  endpoint?: string;
+  runs: number;
+  spend: number;
+}
+
+/** Workspace spend summary aggregated from the full runs list. */
+export interface SpendReport {
+  runs: number;
+  spend: number;
+  currency: string;
+  firstRunAt?: string;
+  lastRunAt?: string;
+  providers: SpendBucket[];
+  topEndpoints: SpendBucket[];
+  topRuns: RunListItem[];
+  /** Runs that FAILED, timed out, or returned a non-2xx provider status. */
+  failedRuns: number;
+  /** Total billed across failed runs ($0 when billing is healthy). */
+  failedSpend: number;
+}
+
 // --- Balance ---
 
 export interface BalanceResponse {
diff --git a/src/commands/spend.ts b/src/commands/spend.ts
new file mode 100644
index 0000000..b18fc04
--- /dev/null
+++ b/src/commands/spend.ts
@@ -0,0 +1,146 @@
+import { Command } from '@cliffy/command';
+import { MonidAPI } from '../api/client.js';
+import type { RunListItem, SpendBucket, SpendReport } from '../api/types.js';
+import { ConfigManager } from '../config/manager.js';
+import { handleError, MonidError } from '../utils/error.js';
+import { printUpdateNotice, applyUpdateNote } from '../utils/update-check.js';
+import { formatSpendReport } from '../output/format.js';
+import { startSpinner, succeedSpinner, stopSpinner, updateSpinner } from '../output/spinner.js';
+
+const PAGE_SIZE = 100;
+const TOP_ENDPOINTS = 5;
+const TOP_RUNS = 3;
+
+export const spendCommand = new Command()
+  .name('spend')
+  .description('Summarize workspace spend by provider and endpoint.')
+  .option('-j, --json', 'Output as JSON.')
+  .action(async ({ json }) => {
+    try {
+      const config = new ConfigManager();
+      const active = config.getActiveKey();
+      if (!active) {
+        throw new MonidError(
+          'AUTH_FAILED',
+          'No active API key. Run "monid keys add" to configure one.',
+        );
+      }
+
+      const api = new MonidAPI({ apiKey: active.credential.key });
+
+      if (!json) {
+        startSpinner('Fetching runs...');
+      }
+
+      const items: RunListItem[] = [];
+      let cursor: string | undefined;
+      do {
+        const page = await api.listRuns(PAGE_SIZE, cursor);
+        items.push(...page.items);
+        cursor = page.cursor ?? undefined;
+        if (!json) updateSpinner(`Fetching runs... ${items.length}`);
+      } while (cursor);
+
+      const report = buildSpendReport(items);
+      const updateInfo = await config.getUpdateInfo();
+
+      if (json) {
+        const output = updateInfo ? applyUpdateNote(report, updateInfo) : report;
+        console.log(JSON.stringify(output, null, 2));
+      } else {
+        succeedSpinner(`Analyzed ${items.length} run(s)`);
+        formatSpendReport(report);
+        if (updateInfo) printUpdateNotice(updateInfo);
+      }
+    } catch (err) {
+      stopSpinner();
+      handleError(err, json);
+    }
+  });
+
+/**
+ * Aggregate the full runs list into a spend report. Items are expected
+ * newest-first, as `/v1/runs` returns them.
+ */
+export function buildSpendReport(items: RunListItem[]): SpendReport {
+  const providers = new Map<string, SpendBucket>();
+  const endpoints = new Map<string, SpendBucket>();
+  let spend = 0;
+  let currency: string | undefined;
+  let failedRuns = 0;
+  let failedSpend = 0;
+
+  for (const run of items) {
+    const cost = run.cost?.value ?? 0;
+    spend += cost;
+    if (currency === undefined && run.cost) currency = run.cost.currency;
+
+    const providerEntry = upsertBucket(providers, run.provider, {
+      provider: run.provider,
+      providerName: run.providerName,
+      runs: 0,
+      spend: 0,
+    });
+    addTo(providerEntry, cost);
+
+    const endpointEntry = upsertBucket(endpoints, `${run.provider} ${run.endpoint}`, {
+      provider: run.provider,
+      providerName: run.providerName,
+      endpoint: run.endpoint,
+      runs: 0,
+      spend: 0,
+    });
+    addTo(endpointEntry, cost);
+
+    if (isFailedRun(run)) {
+      failedRuns += 1;
+      failedSpend += cost;
+    }
+  }
+
+  const topRuns = [...items]
+    .sort((a, b) => (b.cost?.value ?? 0) - (a.cost?.value ?? 0))
+    .slice(0, TOP_RUNS);
+
+  return {
+    runs: items.length,
+    spend,
+    currency: currency ?? 'USD',
+    providers: bySpend(providers),
+    topEndpoints: bySpend(endpoints).slice(0, TOP_ENDPOINTS),
+    topRuns,
+    failedRuns,
+    failedSpend,
+    firstRunAt: items[items.length - 1]?.createdAt,
+    lastRunAt: items[0]?.createdAt,
+  };
+}
+
+/** A run that failed outright: errored, timed out, or the provider itself
+ *  returned a non-2xx status. Excludes BLOCKED runs — those are a budget
+ *  guardrail firing as intended, not a failure. */
+function isFailedRun(run: RunListItem): boolean {
+  if (run.status === 'FAILED' || run.status === 'TIME_OUT') return true;
+  const httpStatus = run.providerResponse?.httpStatus;
+  return httpStatus !== undefined && (httpStatus < 200 || httpStatus >= 300);
+}
+
+function upsertBucket(
+  map: Map<string, SpendBucket>,
+  key: string,
+  seed: SpendBucket,
+): SpendBucket {
+  const existing = map.get(key);
+  if (existing) return existing;
+  map.set(key, seed);
+  return seed;
+}
+
+function addTo(entry: SpendBucket, cost: number): void {
+  entry.runs += 1;
+  entry.spend += cost;
+}
+
+function bySpend(map: Map<string, SpendBucket>): SpendBucket[] {
+  return [...map.values()].sort((a, b) => b.spend - a.spend);
+}
diff --git a/src/index.ts b/src/index.ts
index a63f410..62241c8 100644
--- a/src/index.ts
+++ b/src/index.ts
@@ -1,6 +1,7 @@
 import { Command } from '@cliffy/command';
 import { VERSION } from './config/constants.js';
 import { balanceCommand } from './commands/balance.js';
+import { spendCommand } from './commands/spend.js';
 import { keysCommand } from './commands/keys/index.js';
 import { discoverCommand } from './commands/discover.js';
 import { inspectCommand } from './commands/inspect.js';
@@ -27,6 +28,7 @@ const cli = new Command()
   .command('whoami', whoamiCommand)
   .command('setup', setupCommand)
   .command('balance', balanceCommand)
+  .command('spend', spendCommand)
   .command('keys', keysCommand);
 
 await cli.parse();
diff --git a/src/output/format.ts b/src/output/format.ts
index adde65f..91353b1 100644
--- a/src/output/format.ts
+++ b/src/output/format.ts
@@ -24,6 +24,7 @@ import type {
   RunError,
   RunDetailResponse,
   RunsListResponse,
+  SpendReport,
   WhoamiResponse,
 } from '../api/types.js';
 import { resourceStateBadge } from './colors.js';
@@ -404,6 +405,67 @@ export function formatBalance(data: BalanceResponse): void {
   console.log();
 }
 
+// --- Spend ---
+
+export function formatSpendReport(report: SpendReport): void {
+  if (report.runs === 0) {
+    console.log(chalk.gray('No runs found.'));
+    return;
+  }
+
+  const range =
+    report.firstRunAt && report.lastRunAt
+      ? chalk.gray(` · ${formatDate(report.firstRunAt)} — ${formatDate(report.lastRunAt)}`)
+      : '';
+  console.log();
+  console.log(
+    `  Total spent: ${chalk.green(`$${report.spend.toFixed(2)}`)} ${report.currency}` + range,
+  );
+  console.log();
+
+  console.log(chalk.bold('By provider'));
+  renderTable(
+    ['Provider', 'Runs', 'Spend'],
+    report.providers.map((b) => [
+      b.providerName || b.provider,
+      String(b.runs),
+      `$${b.spend.toFixed(4)}`,
+    ]),
+    { columns: { 1: { align: 'right' }, 2: { align: 'right' } } },
+  );
+
+  console.log();
+  console.log(chalk.bold('Top endpoints'));
+  renderTable(
+    ['Provider', 'Endpoint', 'Runs', 'Spend'],
+    report.topEndpoints.map((b) => [
+      b.providerName || b.provider,
+      b.endpoint ?? '-',
+      String(b.runs),
+      `$${b.spend.toFixed(4)}`,
+    ]),
+    { columns: { 2: { align: 'right' }, 3: { align: 'right' } } },
+  );
+
+  console.log();
+  console.log(chalk.bold('Top runs'));
+  renderTable(
+    ['Run ID', 'Endpoint', 'Results', 'Cost'],
+    report.topRuns.map((r) => [
+      chalk.bold(r.runId),
+      r.endpoint,
+      r.resultCount !== undefined ? String(r.resultCount) : '-',
+      r.cost ? `$${r.cost.value.toFixed(4)}` : '-',
+    ]),
+    { columns: { 2: { align: 'right' }, 3: { align: 'right' } } },
+  );
+
+  console.log();
+  const failedLine = `Failed runs: ${report.failedRuns} · billed $${report.failedSpend.toFixed(2)}`;
+  console.log(`  ${report.failedSpend > 0 ? chalk.yellow(failedLine) : chalk.gray(failedLine)}`);
+  console.log();
+}
+
 // --- Whoami ---
 
 export function formatWhoami(data: WhoamiResponse): void {
diff --git a/test/commands/spend.test.ts b/test/commands/spend.test.ts
new file mode 100644
index 0000000..11545db
--- /dev/null
+++ b/test/commands/spend.test.ts
@@ -0,0 +1,96 @@
+import { describe, it, expect } from 'bun:test';
+import { buildSpendReport } from '../../src/commands/spend.js';
+import type { RunListItem } from '../../src/api/types.js';
+
+function run(overrides: Partial<RunListItem>): RunListItem {
+  return {
+    runId: 'run_1',
+    caller: 'USER#u_1',
+    provider: 'tikhub',
+    providerName: 'TikHub',
+    endpoint: '/a',
+    status: 'COMPLETED',
+    providerResponse: { httpStatus: 200 },
+    price: { type: 'PER_CALL', amount: { value: 0.0015, currency: 'USD' } },
+    cost: { value: 0.0015, currency: 'USD' },
+    createdAt: '2026-08-24T00:00:00.000Z',
+    ...overrides,
+  };
+}
+
+describe('buildSpendReport', () => {
+  it('totals spend and groups by provider and endpoint', () => {
+    const report = buildSpendReport([
+      run({ runId: 'run_1', endpoint: '/a', cost: { value: 0.001, currency: 'USD' } }),
+      run({ runId: 'run_2', endpoint: '/b', cost: { value: 0.002, currency: 'USD' } }),
+      run({
+        runId: 'run_3',
+        provider: 'apify',
+        providerName: 'Apify',
+        endpoint: '/c',
+        cost: { value: 0.5, currency: 'USD' },
+      }),
+    ]);
+
+    expect(report.runs).toBe(3);
+    expect(report.spend).toBeCloseTo(0.503);
+    expect(report.providers.map((b) => b.provider)).toEqual(['apify', 'tikhub']);
+    expect(report.providers[0].spend).toBeCloseTo(0.5);
+    expect(report.providers[1].runs).toBe(2);
+    expect(report.topEndpoints[0]).toMatchObject({ provider: 'apify', endpoint: '/c' });
+  });
+
+  it('ranks top runs by cost', () => {
+    const report = buildSpendReport([
+      run({ runId: 'run_cheap', cost: { value: 0.001, currency: 'USD' } }),
+      run({ runId: 'run_big', cost: { value: 1, currency: 'USD' } }),
+      run({ runId: 'run_free', cost: null }),
+    ]);
+
+    expect(report.topRuns[0].runId).toBe('run_big');
+    expect(report.topRuns).toHaveLength(3);
+  });
+
+  it('counts failed runs and their billed spend', () => {
+    const report = buildSpendReport([
+      run({}),
+      run({ runId: 'run_400', providerResponse: { httpStatus: 400 }, cost: null }),
+      run({ runId: 'run_timeout', status: 'TIME_OUT', cost: null }),
+      run({
+        runId: 'run_billed_fail',
+        providerResponse: { httpStatus: 500 },
+        cost: { value: 0.01, currency: 'USD' },
+      }),
+    ]);
+
+    expect(report.failedRuns).toBe(3);
+    expect(report.failedSpend).toBeCloseTo(0.01);
+  });
+
+  it('does not count blocked runs as failures', () => {
+    const report = buildSpendReport([
+      run({ runId: 'run_blocked', status: 'BLOCKED', providerResponse: undefined, cost: null }),
+    ]);
+
+    expect(report.failedRuns).toBe(0);
+  });
+
+  it('reads the date range from newest-first items', () => {
+    const report = buildSpendReport([
+      run({ runId: 'run_new', createdAt: '2026-08-24T00:00:00.000Z' }),
+      run({ runId: 'run_old', createdAt: '2026-07-21T00:00:00.000Z' }),
+    ]);
+
+    expect(report.firstRunAt).toBe('2026-07-21T00:00:00.000Z');
+    expect(report.lastRunAt).toBe('2026-08-24T00:00:00.000Z');
+  });
+
+  it('returns an empty report for no runs', () => {
+    const report = buildSpendReport([]);
+
+    expect(report.runs).toBe(0);
+    expect(report.spend).toBe(0);
+    expect(report.providers).toEqual([]);
+    expect(report.firstRunAt).toBeUndefined();
+  });
+});