diff --git a/src/backend/drivers/ai-speech2txt/XAISpeechToTextDriver.test.ts b/src/backend/drivers/ai-speech2txt/XAISpeechToTextDriver.test.ts
new file mode 100644
index 0000000000..be4551430d
--- /dev/null
+++ b/src/backend/drivers/ai-speech2txt/XAISpeechToTextDriver.test.ts
@@ -0,0 +1,616 @@
+/*
+ * Copyright (C) 2024-present Puter Technologies Inc.
+ *
+ * This file is part of Puter.
+ *
+ * Puter is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as published
+ * by the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
+/**
+ * Offline unit tests for XAISpeechToTextDriver.
+ *
+ * Boots a real PuterServer (in-memory sqlite + dynamo + s3 + mock
+ * redis) configured with an xAI API key, then drives
+ * `server.drivers.aiSpeech2TxtXai` directly. xAI has no SDK — the
+ * driver calls the REST `/v1/stt` endpoint via `fetch` — so global
+ * `fetch` is spied for each request shape assertion. Audio inputs use
+ * `data:` URLs through the live `loadFileInput`, and the FS-resolution
+ * branch is exercised by writing a real file via `server.services.fs.write`.
+ */
+
+import {
+    afterAll,
+    afterEach,
+    beforeAll,
+    beforeEach,
+    describe,
+    expect,
+    it,
+    vi,
+    type MockInstance,
+} from 'vitest';
+import { v4 as uuidv4 } from 'uuid';
+
+import type { Actor } from '../../core/actor.js';
+import { runWithContext } from '../../core/context.js';
+import { PuterServer } from '../../server.js';
+import type { MeteringService } from '../../services/metering/MeteringService.js';
+import { setupTestServer } from '../../testUtil.js';
+import { generateDefaultFsentries } from '../../util/userProvisioning.js';
+import type { XAISpeechToTextDriver } from './XAISpeechToTextDriver.js';
+
+// Mirror the driver's per-second cost so test expectations stay in lockstep
+// with the real value rather than restating an arbitrary literal.
+const UCENTS_PER_SECOND = 2778;
+
+// ── Test harness ────────────────────────────────────────────────────
+
+let server: PuterServer;
+let driver: XAISpeechToTextDriver;
+let fetchSpy: MockInstance<typeof fetch>;
+let hasCreditsSpy: MockInstance<MeteringService['hasEnoughCredits']>;
+let incrementUsageSpy: MockInstance<MeteringService['incrementUsage']>;
+
+beforeAll(async () => {
+    server = await setupTestServer({
+        providers: {
+            xai: { apiKey: 'xai-test-key' },
+        },
+    } as never);
+    driver = server.drivers.aiSpeech2TxtXai as unknown as XAISpeechToTextDriver;
+});
+
+afterAll(async () => {
+    await server?.shutdown();
+});
+
+beforeEach(() => {
+    fetchSpy = vi.spyOn(globalThis, 'fetch') as MockInstance<typeof fetch>;
+    hasCreditsSpy = vi.spyOn(server.services.metering, 'hasEnoughCredits');
+    incrementUsageSpy = vi.spyOn(server.services.metering, 'incrementUsage');
+});
+
+afterEach(() => {
+    vi.restoreAllMocks();
+});
+
+const makeUser = async (): Promise<{ actor: Actor; userId: number }> => {
+    const username = `xai-stt-${Math.random().toString(36).slice(2, 10)}`;
+    const created = await server.stores.user.create({
+        username,
+        uuid: uuidv4(),
+        password: null,
+        email: `${username}@test.local`,
+        free_storage: 100 * 1024 * 1024,
+        requires_email_confirmation: false,
+    });
+    await generateDefaultFsentries(
+        server.clients.db,
+        server.stores.user,
+        created,
+    );
+    const refreshed = (await server.stores.user.getById(created.id))!;
+    return {
+        userId: refreshed.id,
+        actor: {
+            user: {
+                id: refreshed.id,
+                uuid: refreshed.uuid,
+                username: refreshed.username,
+                email: refreshed.email ?? null,
+                email_confirmed: true,
+            } as Actor['user'],
+        },
+    };
+};
+
+const withActor = <T>(actor: Actor, fn: () => T | Promise<T>): Promise<T> =>
+    Promise.resolve(runWithContext({ actor }, fn));
+
+const dataUrl = (buffer: Buffer, mime: string) =>
+    `data:${mime};base64,${buffer.toString('base64')}`;
+
+const sttResponse = (body: Record<string, unknown>, status = 200) =>
+    new Response(JSON.stringify(body), {
+        status,
+        headers: { 'content-type': 'application/json' },
+    });
+
+// ── getReportedCosts ────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver.getReportedCosts', () => {
+    it('reports a single per-second line item at the documented xAI rate', () => {
+        const reported = driver.getReportedCosts();
+        expect(reported).toEqual([
+            {
+                usageType: 'xai:stt:second',
+                ucentsPerUnit: UCENTS_PER_SECOND,
+                unit: 'second',
+                source: 'driver:aiSpeech2Txt/xai',
+            },
+        ]);
+    });
+});
+
+// ── list_models ─────────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver.list_models', () => {
+    it('returns the single xai-stt entry with diarization support', async () => {
+        const models = await driver.list_models();
+        expect(models).toHaveLength(1);
+        expect(models[0]).toEqual({
+            id: 'xai-stt',
+            name: 'xAI Speech to Text',
+            type: 'transcription',
+            response_formats: ['json'],
+            supports_prompt: false,
+            supports_logprobs: false,
+            supports_diarization: true,
+        });
+    });
+});
+
+// ── test_mode bypass ────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver test_mode', () => {
+    it('returns the canned sample for transcribe, bypassing all I/O', async () => {
+        const result = (await driver.transcribe({
+            file: undefined,
+            test_mode: true,
+        })) as { text: string; model: string; words: unknown[] };
+
+        expect(result.text).toMatch(/sample transcription/i);
+        expect(result.model).toBe('xai-stt');
+        expect(Array.isArray(result.words)).toBe(true);
+        // No file required, no actor required, no fetch / metering hit.
+        expect(fetchSpy).not.toHaveBeenCalled();
+        expect(hasCreditsSpy).not.toHaveBeenCalled();
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+
+    it('returns the canned sample for translate (same alias, no fetch)', async () => {
+        const result = (await driver.translate({
+            file: undefined,
+            test_mode: true,
+        })) as { model: string };
+        expect(result.model).toBe('xai-stt');
+        expect(fetchSpy).not.toHaveBeenCalled();
+    });
+});
+
+// ── Argument validation ─────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver argument validation', () => {
+    it('throws 400 when file is missing', async () => {
+        const { actor } = await makeUser();
+        await expect(
+            withActor(actor, () => driver.transcribe({ file: undefined })),
+        ).rejects.toMatchObject({ statusCode: 400 });
+        expect(fetchSpy).not.toHaveBeenCalled();
+    });
+
+    it('throws 401 when no actor is on the request context', async () => {
+        await expect(
+            driver.transcribe({
+                file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+            }),
+        ).rejects.toMatchObject({ statusCode: 401 });
+        expect(fetchSpy).not.toHaveBeenCalled();
+    });
+});
+
+// ── Missing API key ─────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver missing API key', () => {
+    it('throws 500 internal_error when no xAI key is configured', async () => {
+        // Boot a separate server WITHOUT an xAI provider entry so the
+        // driver leaves its key unset. test_mode and validation paths
+        // are unaffected — only the network branch should reject.
+        const bareServer = await setupTestServer();
+        try {
+            const bareDriver =
+                bareServer.drivers.aiSpeech2TxtXai as unknown as XAISpeechToTextDriver;
+            await expect(
+                bareDriver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ).rejects.toMatchObject({
+                statusCode: 500,
+                legacyCode: 'internal_error',
+            });
+        } finally {
+            await bareServer.shutdown();
+        }
+    });
+});
+
+// ── Credit gate ─────────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver credit gate', () => {
+    it('throws 402 BEFORE hitting xAI when actor lacks credits', async () => {
+        hasCreditsSpy.mockResolvedValueOnce(false);
+        const { actor } = await makeUser();
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('audio-bytes'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toMatchObject({ statusCode: 402 });
+
+        expect(fetchSpy).not.toHaveBeenCalled();
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+
+    it('asks hasEnoughCredits for estimated seconds × per-second ucents', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        // 32000 bytes / 16000 bytes-per-second ⇒ 2s estimated.
+        const audio = Buffer.alloc(32000, 0);
+        await withActor(actor, () =>
+            driver.transcribe({ file: dataUrl(audio, 'audio/mp3') }),
+        );
+
+        expect(hasCreditsSpy.mock.calls[0]![1]).toBe(UCENTS_PER_SECOND * 2);
+    });
+
+    it('estimates 60 seconds for URL inputs (we cannot inspect size locally)', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        await withActor(actor, () =>
+            driver.transcribe({ file: 'https://example.com/clip.mp3' }),
+        );
+
+        expect(hasCreditsSpy.mock.calls[0]![1]).toBe(UCENTS_PER_SECOND * 60);
+    });
+});
+
+// ── Audio input handling ────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver audio input handling', () => {
+    it('decodes a base64 data URL and POSTs it as the multipart `file` field', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'hello world', duration: 1 }),
+        );
+
+        const audioBytes = Buffer.from('fake-mp3-bytes');
+        await withActor(actor, () =>
+            driver.transcribe({ file: dataUrl(audioBytes, 'audio/mp3') }),
+        );
+
+        const [url, init] = fetchSpy.mock.calls[0]!;
+        expect(String(url)).toBe('https://api.x.ai/v1/stt');
+        const initObj = init as RequestInit;
+        expect(initObj.method).toBe('POST');
+        expect((initObj.headers as Record<string, string>).Authorization).toBe(
+            'Bearer xai-test-key',
+        );
+        // multipart body — `file` is present, `url` is not.
+        const form = initObj.body as FormData;
+        expect(form).toBeInstanceOf(FormData);
+        expect(form.get('url')).toBeNull();
+        const filePart = form.get('file');
+        expect(filePart).toBeInstanceOf(Blob);
+        const blob = filePart as Blob;
+        expect(blob.type).toBe('audio/mp3');
+        const sent = Buffer.from(await blob.arrayBuffer());
+        expect(sent.equals(audioBytes)).toBe(true);
+    });
+
+    it('forwards an HTTP(S) URL as the `url` field and skips local FS read', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        await withActor(actor, () =>
+            driver.transcribe({ file: 'https://example.com/audio.mp3' }),
+        );
+
+        const form = (fetchSpy.mock.calls[0]![1] as RequestInit)
+            .body as FormData;
+        expect(form.get('url')).toBe('https://example.com/audio.mp3');
+        expect(form.get('file')).toBeNull();
+    });
+
+    it('resolves an FS path through the live FSService and preserves filename/mime', async () => {
+        const { actor, userId } = await makeUser();
+        const audioBytes = Buffer.from('fs-backed-audio-data');
+        await server.services.fs.write(userId, {
+            fileMetadata: {
+                path: `/${actor.user.username}/clip.mp3`,
+                size: audioBytes.byteLength,
+                contentType: 'audio/mpeg',
+            },
+            fileContent: audioBytes,
+        });
+
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        await withActor(actor, () =>
+            driver.transcribe({
+                file: { path: `/${actor.user.username}/clip.mp3` },
+            }),
+        );
+
+        const form = (fetchSpy.mock.calls[0]![1] as RequestInit)
+            .body as FormData;
+        const filePart = form.get('file') as File;
+        expect(filePart).toBeInstanceOf(Blob);
+        expect(filePart.type).toBe('audio/mpeg');
+        // FormData.append(name, blob, filename) — Blob becomes a File with .name.
+        expect(filePart.name).toBe('clip.mp3');
+    });
+});
+
+// ── Request shape ───────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver request shape', () => {
+    it('forwards language / format / diarize / multichannel / channels / audio_format / sample_rate', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        await withActor(actor, () =>
+            driver.transcribe({
+                file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                language: 'en',
+                format: true,
+                diarize: true,
+                multichannel: true,
+                channels: 2,
+                audio_format: 'pcm',
+                sample_rate: 16000,
+            }),
+        );
+
+        const form = (fetchSpy.mock.calls[0]![1] as RequestInit)
+            .body as FormData;
+        expect(form.get('language')).toBe('en');
+        expect(form.get('format')).toBe('true');
+        expect(form.get('diarize')).toBe('true');
+        expect(form.get('multichannel')).toBe('true');
+        expect(form.get('channels')).toBe('2');
+        expect(form.get('audio_format')).toBe('pcm');
+        expect(form.get('sample_rate')).toBe('16000');
+    });
+
+    it('omits optional fields when the caller does not supply them', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        await withActor(actor, () =>
+            driver.transcribe({
+                file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+            }),
+        );
+
+        const form = (fetchSpy.mock.calls[0]![1] as RequestInit)
+            .body as FormData;
+        expect(form.get('language')).toBeNull();
+        expect(form.get('format')).toBeNull();
+        expect(form.get('diarize')).toBeNull();
+        expect(form.get('multichannel')).toBeNull();
+        expect(form.get('channels')).toBeNull();
+        expect(form.get('audio_format')).toBeNull();
+        expect(form.get('sample_rate')).toBeNull();
+    });
+
+    it('routes translate() to the same /v1/stt endpoint as transcribe()', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 1 }),
+        );
+
+        await withActor(actor, () =>
+            driver.translate({
+                file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+            }),
+        );
+
+        expect(String(fetchSpy.mock.calls[0]![0])).toBe('https://api.x.ai/v1/stt');
+    });
+});
+
+// ── Response shape ──────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver response shape', () => {
+    it('forwards the parsed xAI JSON response verbatim', async () => {
+        const { actor } = await makeUser();
+        const upstream = {
+            text: 'hello world',
+            language: 'English',
+            duration: 2.5,
+            words: [{ text: 'hello', start: 0, end: 1 }],
+        };
+        fetchSpy.mockResolvedValueOnce(sttResponse(upstream));
+
+        const result = await withActor(actor, () =>
+            driver.transcribe({
+                file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+            }),
+        );
+
+        expect(result).toEqual(upstream);
+    });
+});
+
+// ── Metering ────────────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver metering', () => {
+    it('meters ceil(duration) seconds × per-second ucents using the API duration', async () => {
+        const { actor } = await makeUser();
+        // Upstream reports 3.2s → driver should ceil to 4s.
+        fetchSpy.mockResolvedValueOnce(
+            sttResponse({ text: 'ok', duration: 3.2 }),
+        );
+
+        await withActor(actor, () =>
+            driver.transcribe({
+                file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+            }),
+        );
+
+        const sttCalls = incrementUsageSpy.mock.calls.filter(
+            ([, type]) => type === 'xai:stt:second',
+        );
+        expect(sttCalls).toHaveLength(1);
+        const [actorArg, , count, cost] = sttCalls[0]!;
+        expect((actorArg as Actor).user.id).toBe(actor.user.id);
+        expect(count).toBe(4);
+        expect(cost).toBe(UCENTS_PER_SECOND * 4);
+    });
+
+    it('falls back to the byte-based estimate when the API omits duration', async () => {
+        const { actor } = await makeUser();
+        // 32000 bytes → estimated 2s; upstream omits `duration`.
+        fetchSpy.mockResolvedValueOnce(sttResponse({ text: 'ok' }));
+
+        await withActor(actor, () =>
+            driver.transcribe({
+                file: dataUrl(Buffer.alloc(32000, 0), 'audio/mp3'),
+            }),
+        );
+
+        const sttCalls = incrementUsageSpy.mock.calls.filter(
+            ([, type]) => type === 'xai:stt:second',
+        );
+        expect(sttCalls).toHaveLength(1);
+        const [, , count, cost] = sttCalls[0]!;
+        expect(count).toBe(2);
+        expect(cost).toBe(UCENTS_PER_SECOND * 2);
+    });
+});
+
+// ── Error paths ─────────────────────────────────────────────────────
+
+describe('XAISpeechToTextDriver error paths', () => {
+    it('maps upstream 4xx to HttpError 400 upstream_bad_request and skips metering', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            new Response('bad request', { status: 400 }),
+        );
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toMatchObject({
+            statusCode: 400,
+            legacyCode: 'upstream_bad_request',
+        });
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+
+    it('maps upstream 5xx to HttpError 400 upstream_provider_unavailable', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(new Response('oops', { status: 503 }));
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toMatchObject({
+            statusCode: 400,
+            legacyCode: 'upstream_provider_unavailable',
+        });
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+
+    it('maps upstream 401/403 to HttpError 500 upstream_auth_failed', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            new Response('forbidden', { status: 403 }),
+        );
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toMatchObject({
+            statusCode: 500,
+            legacyCode: 'upstream_auth_failed',
+        });
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+
+    it('maps upstream 429 to HttpError 429 upstream_rate_limited', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            new Response('slow down', { status: 429 }),
+        );
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toMatchObject({
+            statusCode: 429,
+            legacyCode: 'upstream_rate_limited',
+        });
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+
+    it('tags upstream errors with provider=xai and the original status', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockResolvedValueOnce(
+            new Response('boom', { status: 502 }),
+        );
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toMatchObject({
+            fields: { provider: 'xai', upstreamStatus: 502 },
+        });
+    });
+
+    it('lets fetch network errors bubble so the driver boundary can decide', async () => {
+        const { actor } = await makeUser();
+        fetchSpy.mockRejectedValueOnce(new Error('connection reset'));
+
+        await expect(
+            withActor(actor, () =>
+                driver.transcribe({
+                    file: dataUrl(Buffer.from('a'), 'audio/mp3'),
+                }),
+            ),
+        ).rejects.toThrow('connection reset');
+        expect(incrementUsageSpy).not.toHaveBeenCalled();
+    });
+});