@mrgoonie/multix 0.15.0 → 0.15.1

package/CHANGELOG.md

@@ -5,6 +5,13 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.15.1](https://github.com/mrgoonie/multix-cli/compare/v0.15.0...v0.15.1) (2026-10-11)
+
+
+### Performance
+
+* **gemini:** stream file uploads with resumable Files API ([#85](https://github.com/mrgoonie/multix-cli/issues/85)) ([b2cc886](https://github.com/mrgoonie/multix-cli/commit/b2cc886cf5bdf960e88aeea5155ae2623547d918))
+
 ## [0.15.0](https://github.com/mrgoonie/multix-cli/compare/v0.14.1...v0.15.0) (2026-10-11)
 
 

package/README.md

@@ -99,6 +99,7 @@ Set at least one provider key. Add to `.env` in your project root or `~/.multix/
 | `CLOUDFLARE_AI_TTS_MODEL` | No | Speech model setting; only `@cf/myshell-ai/melotts` is accepted |
 | `MULTIX_OUTPUT_DIR` | No | Override default output dir (`./multix-output`) |
 | `MULTIX_MAX_DOWNLOAD_MB` | No | Size limit per downloaded Replicate output file (default `2048`) |
+| `MULTIX_UPLOAD_MIN_KBPS` | No | Slowest Gemini Files API upload speed to allow for, in KB/s (default `1000`); lower it on slow links |
 | `MULTIX_IMAGE_FORMAT` | No | `webp` (default) or `original` to keep provider image format/quality |
 | `OPENROUTER_IMAGE_MODEL` | No | Default OpenRouter model |
 | `OPENROUTER_FALLBACK_MODELS` | No | Comma-separated fallback model ids |

package/dist/cli.js

@@ -46,8 +46,8 @@ var init_errors = __esm({
     "use strict";
     init_redact();
     MultixError = class extends Error {
-      constructor(message, code) {
-        super(message);
+      constructor(message, code, options) {
+        super(message, options);
         this.code = code;
         this.name = "MultixError";
         Error.captureStackTrace?.(
@@ -64,8 +64,8 @@ var init_errors = __esm({
       }
     };
     ProviderError = class extends MultixError {
-      constructor(message, provider) {
-        super(message, "PROVIDER_ERROR");
+      constructor(message, provider, options) {
+        super(message, "PROVIDER_ERROR", options);
         this.provider = provider;
         this.name = "ProviderError";
       }
@@ -78,8 +78,8 @@ var init_errors = __esm({
       }
     };
     HttpError = class extends MultixError {
-      constructor(status, snippet, url, retryAfterMs) {
-        super(`HTTP ${status} from ${redactUrl(url)}: ${snippet}`, "HTTP_ERROR");
+      constructor(status, snippet, url, retryAfterMs, options) {
+        super(`HTTP ${status} from ${redactUrl(url)}: ${snippet}`, "HTTP_ERROR", options);
         this.status = status;
         this.snippet = snippet;
         this.retryAfterMs = retryAfterMs;
@@ -4651,11 +4651,20 @@ init_errors();
 import fs20 from "fs";
 import os2 from "os";
 import path23 from "path";
+
+// src/core/limits.ts
+var FILES_API_MAX_BYTES = 2e9;
+
+// src/core/stdin-file.ts
 var STDIN_ARG = "-";
 var STDIN_LABEL = "<stdin>";
 var HEAD_BYTES = 4096;
-var MAX_STDIN_BYTES = 2 * 1024 ** 3;
+var MAX_STDIN_BYTES = FILES_API_MAX_BYTES;
 var WAIT_HINT_MS = 2e3;
+function stdinTooLargeMessage(maxBytes) {
+  const bytes = maxBytes.toLocaleString("en-US");
+  return maxBytes === FILES_API_MAX_BYTES ? `Piped stdin exceeds the Gemini Files API limit of 2 GB (${bytes} bytes)` : `Piped stdin exceeds the ${bytes}-byte limit`;
+}
 var ascii = (buf, start, end) => buf.toString("latin1", start, end);
 var HEIF_BRANDS = /* @__PURE__ */ new Set([
   "heic",
@@ -4795,9 +4804,7 @@ async function stdinToTempFile(opts = {}) {
         clearTimeout(hint);
         const buf = typeof chunk === "string" ? Buffer.from(chunk) : chunk;
         if (size + buf.length > maxBytes) {
-          throw new ValidationError(
-            `Piped stdin exceeds ${Math.floor(maxBytes / 1024 ** 2)} MB; save it to a file instead`
-          );
+          throw new ValidationError(stdinTooLargeMessage(maxBytes));
         }
         if (head.length < HEAD_BYTES) {
           head = Buffer.concat([head, buf.subarray(0, HEAD_BYTES - head.length)]);
@@ -4859,9 +4866,9 @@ async function resolveFileInputs(files, opts = {}) {
 }
 
 // src/providers/gemini/client.ts
-init_errors();
 import fs21 from "fs";
 import path25 from "path";
+init_errors();
 
 // src/providers/gemini/key-pool.ts
 init_errors();
@@ -4985,49 +4992,123 @@ function withKey(opts, fn) {
 function authHeaders2(apiKey) {
   return { "x-goog-api-key": apiKey };
 }
+var DEFAULT_TIMEOUT_MS3 = 3e5;
+var UPLOAD_MIN_KBPS_ENV = "MULTIX_UPLOAD_MIN_KBPS";
+var DEFAULT_MIN_UPLOAD_BYTES_PER_SEC = 1e6;
+var UPLOAD_TIMEOUT_SLACK_MS = 6e4;
+function minUploadBytesPerSec() {
+  const raw = resolveKey(UPLOAD_MIN_KBPS_ENV)?.trim();
+  if (!raw) return DEFAULT_MIN_UPLOAD_BYTES_PER_SEC;
+  const kbps = Number(raw);
+  if (!Number.isFinite(kbps) || kbps < 1) {
+    throw new ConfigError(
+      `${UPLOAD_MIN_KBPS_ENV} must be a number of kilobytes per second, at least 1, got "${raw}"`
+    );
+  }
+  return kbps * 1e3;
+}
+function uploadTimeoutMs(size, timeoutMs = DEFAULT_TIMEOUT_MS3, bytesPerSec = DEFAULT_MIN_UPLOAD_BYTES_PER_SEC) {
+  return Math.max(timeoutMs, UPLOAD_TIMEOUT_SLACK_MS + Math.ceil(size / bytesPerSec * 1e3));
+}
 async function uploadFile(filePath, opts = {}) {
-  return withKey(opts, (apiKey) => uploadFileWithKey(filePath, apiKey, opts));
+  const size = fs21.statSync(filePath).size;
+  if (size > FILES_API_MAX_BYTES) {
+    throw new ValidationError(
+      `File too large for the Gemini Files API: ${path25.basename(filePath)} is ${formatMb3(size)} MB (limit 2 GB = ${FILES_API_MAX_BYTES} bytes)`
+    );
+  }
+  const bytesPerSec = minUploadBytesPerSec();
+  return withKey(opts, (apiKey) => uploadFileWithKey(filePath, apiKey, { ...opts, bytesPerSec }));
+}
+function formatMb3(bytes) {
+  return (bytes / 1024 / 1024).toFixed(2);
+}
+async function uploadFetch(url, init, timeoutMs, step) {
+  let resp;
+  try {
+    resp = await globalThis.fetch(url, {
+      ...init,
+      // With the default "follow", fetch clones the request body (tee) to replay
+      // it on a redirect, and that copy keeps every chunk in memory.
+      redirect: "error",
+      signal: AbortSignal.timeout(timeoutMs)
+    });
+  } catch (err2) {
+    const e = err2;
+    const timedOut = e?.name === "TimeoutError";
+    const redirected = e?.cause instanceof Error && e.cause.message === "unexpected redirect";
+    const reason = timedOut ? `timed out after ${Math.round(timeoutMs / 1e3)} s` : e?.cause instanceof Error ? `${e.message}: ${e.cause.message}` : String(e?.message ?? err2);
+    const message = `File upload failed: ${reason}`;
+    if (redirected || timedOut && step === "upload") {
+      throw new ProviderError(message, "gemini", { cause: err2 });
+    }
+    throw new HttpError(0, message, FILES_BASE, void 0, { cause: err2 });
+  }
+  if (!resp.ok) {
+    const txt = await resp.text();
+    throw new HttpError(resp.status, `File upload failed: ${txt.slice(0, 500)}`, FILES_BASE);
+  }
+  return resp;
+}
+async function readFinalizedFile(resp) {
+  const status = resp.headers.get("x-goog-upload-status")?.trim().toLowerCase();
+  let file;
+  let readError;
+  try {
+    file = (await resp.json())?.file;
+  } catch (err2) {
+    readError = err2;
+  }
+  if (status !== "final" || typeof file?.uri !== "string" || typeof file?.name !== "string") {
+    const detail = status !== "final" ? `upload status is "${status?.slice(0, 40) ?? "missing"}"` : readError ? "unreadable response body" : "no file in the response";
+    throw new ProviderError(`File upload failed: upload was not finalized (${detail})`, "gemini", {
+      cause: readError
+    });
+  }
+  return file;
 }
 async function uploadFileWithKey(filePath, apiKey, opts) {
-  const { timeoutMs = 3e5, logger } = opts;
+  const { timeoutMs = DEFAULT_TIMEOUT_MS3, logger, bytesPerSec } = opts;
   const mimeType = getMimeType(filePath);
-  const fileBytes = fs21.readFileSync(filePath);
   const displayName = path25.basename(filePath);
-  const boundary = `multix_${Date.now()}`;
-  const metaJson = JSON.stringify({ file: { displayName, mimeType } });
-  const body = Buffer.concat([
-    Buffer.from(
-      `--${boundary}\r
-Content-Type: application/json; charset=UTF-8\r
-\r
-${metaJson}\r
---${boundary}\r
-Content-Type: ${mimeType}\r
-\r
-`
-    ),
-    fileBytes,
-    Buffer.from(`\r
---${boundary}--`)
-  ]);
-  logger?.debug(`Uploading ${displayName} (${(fileBytes.length / 1024 / 1024).toFixed(2)} MB)`);
-  const resp = await globalThis.fetch(FILES_BASE, {
-    method: "POST",
-    headers: {
-      "x-goog-api-key": apiKey,
-      "Content-Type": `multipart/related; boundary=${boundary}`,
-      "Content-Length": String(body.length),
-      "X-Goog-Upload-Protocol": "multipart"
+  const file = await fs21.openAsBlob(filePath);
+  logger?.debug(`Uploading ${displayName} (${formatMb3(file.size)} MB)`);
+  const start = await uploadFetch(
+    FILES_BASE,
+    {
+      method: "POST",
+      headers: {
+        "x-goog-api-key": apiKey,
+        "Content-Type": "application/json",
+        "X-Goog-Upload-Protocol": "resumable",
+        "X-Goog-Upload-Command": "start",
+        "X-Goog-Upload-Header-Content-Length": String(file.size),
+        "X-Goog-Upload-Header-Content-Type": mimeType
+      },
+      body: JSON.stringify({ file: { display_name: displayName } })
     },
-    body,
-    signal: AbortSignal.timeout(timeoutMs)
-  });
-  if (!resp.ok) {
-    const txt = await resp.text();
-    throw new HttpError(resp.status, `File upload failed: ${txt.slice(0, 500)}`, FILES_BASE);
+    timeoutMs,
+    "start"
+  );
+  const uploadUrl = start.headers.get("x-goog-upload-url");
+  await start.body?.cancel();
+  if (!uploadUrl) {
+    throw new ProviderError("File upload failed: the Files API returned no upload URL", "gemini");
   }
-  const data = await resp.json();
-  let fileRef = data.file;
+  const resp = await uploadFetch(
+    uploadUrl,
+    {
+      method: "POST",
+      headers: {
+        "X-Goog-Upload-Offset": "0",
+        "X-Goog-Upload-Command": "upload, finalize"
+      },
+      body: file
+    },
+    uploadTimeoutMs(file.size, timeoutMs, bytesPerSec),
+    "upload"
+  );
+  let fileRef = await readFinalizedFile(resp);
   if (requiresProcessingWait(mimeType)) {
     const deadline = Date.now() + timeoutMs;
     while (fileRef.state === "PROCESSING" && Date.now() < deadline) {
@@ -9474,7 +9555,7 @@ import path45 from "path";
 import fs37 from "fs";
 import path44 from "path";
 init_errors();
-var DEFAULT_TIMEOUT_MS3 = 24e4;
+var DEFAULT_TIMEOUT_MS4 = 24e4;
 function requireOpenAIKey() {
   const key = resolveKey("OPENAI_API_KEY");
   if (!key) {
@@ -9490,7 +9571,7 @@ function endpoint(pathname) {
 function authHeaders4(apiKey, extra) {
   return { Authorization: `Bearer ${apiKey}`, ...extra ?? {} };
 }
-function postOpenAIJson(pathname, body, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS3) {
+function postOpenAIJson(pathname, body, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS4) {
   return httpJson({
     url: endpoint(pathname),
     method: "POST",
@@ -9499,7 +9580,7 @@ function postOpenAIJson(pathname, body, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS3)
     timeoutMs
   });
 }
-async function postOpenAIForm(pathname, form, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS3) {
+async function postOpenAIForm(pathname, form, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS4) {
   const url = endpoint(pathname);
   let res;
   try {
@@ -9518,7 +9599,7 @@ async function postOpenAIForm(pathname, form, apiKey, timeoutMs = DEFAULT_TIMEOU
   }
   return res.json();
 }
-async function postOpenAIBinary(pathname, body, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS3) {
+async function postOpenAIBinary(pathname, body, apiKey, timeoutMs = DEFAULT_TIMEOUT_MS4) {
   const url = endpoint(pathname);
   let res;
   try {
@@ -11441,7 +11522,7 @@ import fs45 from "fs";
 import path52 from "path";
 var DATA_URI_MAX_BYTES = 256 * 1024;
 var UPLOAD_MAX_BYTES = 100 * 1024 * 1024;
-function formatMb3(bytes) {
+function formatMb4(bytes) {
   return `${(bytes / (1024 * 1024)).toFixed(1)}MB`;
 }
 async function resolveFileInput(value, client, logger) {
@@ -11457,7 +11538,7 @@ async function resolveFileInput(value, client, logger) {
   if (stat.size === 0) throw new ValidationError(`Input file is empty: ${value}`);
   if (stat.size >= UPLOAD_MAX_BYTES) {
     throw new ValidationError(
-      `Input file ${value} is ${formatMb3(stat.size)}; Replicate accepts uploads under 100MB. Host it and pass an https URL instead.`
+      `Input file ${value} is ${formatMb4(stat.size)}; Replicate accepts uploads under 100MB. Host it and pass an https URL instead.`
     );
   }
   const filename = path52.basename(filePath);
@@ -11467,7 +11548,7 @@ async function resolveFileInput(value, client, logger) {
     logger.debug(`Inlining ${filename} (${stat.size} bytes) as a data URI`);
     return `data:${contentType};base64,${bytes.toString("base64")}`;
   }
-  logger.info(`Uploading ${filename} (${formatMb3(stat.size)}) to Replicate Files...`);
+  logger.info(`Uploading ${filename} (${formatMb4(stat.size)}) to Replicate Files...`);
   const file = await client.uploadFile({
     data: new Blob([bytes], { type: contentType }),
     filename,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mrgoonie/multix",
-  "version": "0.15.0",
+  "version": "0.15.1",
   "description": "AI multimodal CLI — OpenAI, Gemini, MiniMax, OpenRouter, Leonardo, BytePlus, ElevenLabs, fal.ai, Atlas Cloud, Replicate, HeyGen, ffmpeg, ImageMagick, doc-to-md",
   "type": "module",
   "bin": {

package/skill/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: multix
 description: AI multimodal CLI — generate and edit images (OpenAI, Codex driver, Gemini Nano Banana, Imagen, MiniMax, OpenRouter, Leonardo, BytePlus Seedream, Cloudflare Workers AI FLUX.1 Schnell, fal.ai FLUX, Replicate FLUX), generate video (Gemini Veo, MiniMax Hailuo, OpenRouter via Veo/Seedance/Kling, Leonardo via Motion/Veo/Kling/Hailuo/Seedance/LTX, BytePlus Seedance, Cloudflare AI Gateway via Replicate, fal.ai queue models, Replicate direct API models incl. P-Video, HeyGen Video Agent and scripted avatar videos), video translation and lipsync (HeyGen), TTS/STT/music (OpenAI, Gemini Flash TTS incl. Gemini 3.8 with delivery-style control, MiniMax, Cloudflare Workers AI MeloTTS, ElevenLabs), 3D models (BytePlus Hyper3D/Hitem3d), analyze/transcribe media, convert documents to Markdown, optimize media via ffmpeg/ImageMagick, and self-update the CLI. Use whenever the user wants to create, edit, or transform images/video/audio/3D/documents from a CLI, mentions any of: OpenAI, Codex image driver, image-to-image, i2i, image edit, watercolor/cyberpunk style transfer, reference image, OpenRouter, Nano Banana, Flux, Seedream, Veo, Hailuo, Seedance, Kling, Cloudflare Workers AI, AI Gateway, fal.ai, Atlas Cloud, Replicate, replicate.com models, Replicate predictions, HeyGen, avatar video, talking head, Video Agent, video translation, dubbing, lipsync, video generation, text-to-video, image-to-video, i2v, ElevenLabs voice cloning, updating/upgrading multix, or asks for batch media optimization.
-version: 0.15.0 # x-release-please-version
+version: 0.15.1 # x-release-please-version
 ---
 
 # multix — AI Multimodal CLI
@@ -387,6 +387,7 @@ Requires `cwebp` (libwebp: `brew install webp`, `apt install webp`) on `PATH`.
 |---|---|
 | `MULTIX_OUTPUT_DIR` | Override default output dir |
 | `MULTIX_MAX_DOWNLOAD_MB` | Size limit per downloaded Replicate output file (default 2048) |
+| `MULTIX_UPLOAD_MIN_KBPS` | Slowest Gemini Files API upload speed to allow for, in KB/s (default 1000); lower it on slow links |
 | `OPENAI_IMAGE_MODEL` / `OPENAI_TTS_MODEL` / `OPENAI_STT_MODEL` | OpenAI model defaults |
 | `OPENROUTER_IMAGE_MODEL` | Default OpenRouter image model |
 | `OPENROUTER_FALLBACK_MODELS` | CSV; applies to `generate` AND `i2i` |