Deploy Arabic Audio Reader worker

.export-manifest.json

@@ -3,13 +3,14 @@
     "Dockerfile": "eb8b1b840c8303bbcc2354a0e60896cf96593ac6828df6179877b19022e97c1c",
     "api/index.py": "b0fd5c43eadd241aea79131d12ea40fe032a97f06410ce1b607e81c45f33d6f2",
     "app/__init__.py": "7eb70257593da06f682a3ddda54a9d260d4fc514f645237f5ca74b08f8da61a6",
-    "app/main.py": "19c6b13c9393871d88a0fef05d8fc0f19d14bef88205ffc229bce3e3a522f0e4",
+    "app/main.py": "0b280a15a15a82c22fc1b1ed85a08a54fb647325919a5e37100d3525e177417f",
     "docs/best-free-arabic-pdf-audio-stack.md": "08234106caacc0207f404b11023656cdc39525b28fedf526e97369edf926c48f",
-    "docs/father-user-guide.md": "a05534fa8ecc4bee94704b6691947ac189f6767a95fd12eb65ae27c4ede1182f",
+    "docs/father-user-guide.md": "2adb2a56c862df395cd77bf80501a937efc7ad72b30455250a97b4193f7003aa",
     "docs/huggingface-model-metadata.md": "4590229078c2048b184787e85e5a00dd687ef5fc90e8d8d0af32538b38363cc2",
     "docs/live-deployment-checklist.md": "7fd21a9316c1d018e2bec0620defcaaca2a690f109e51b5902c7d157244834ac",
-    "docs/ocr-readability-benchmark.md": "f93f09729f5e8bd5f938afad9490b471452ca549d081ff7700161cc1dc961453",
-    "docs/production-worker-architecture.md": "1264c16b83948385026aca0fab18e7963fa5056a178fa381380659352274b4ff",
+    "docs/ocr-and-voice-ranking.md": "092406ab8c494a160215adea223bbebd364bc7b546c63ed7db0e840ac9318412",
+    "docs/ocr-readability-benchmark.md": "12be9cc094e56454666ff71cc0bf9d14f39cbc7acb4c1a49abf10ffd2e4ef7df",
+    "docs/production-worker-architecture.md": "bfe77f5d718cbf829a30863277ed2e8af1ba227d13067700cdc75b02fe5245c4",
     "docs/recommended-decision-card.json": "97e4607db20ac19cadc9b894d6406517bcb37f8ccc6ecbe6c0c41f5f2463398e",
     "docs/recommended-decision-card.md": "f69bbe66d7977a4877f934212862159495ea5a4547997e059f5c4e1b8d6d6cb9",
     "docs/recommended-free-stack.md": "6156deac80f5656ff4cd33d726061965b6e2a6fbc8db4ee4123b2b43e42aa40d",
@@ -30,7 +31,7 @@
     "scripts/arabic_qwen_ocr_extract.py": "485e9f3cdf2ced92c666b2f483d4aa37a65cb34052a4967beac7183d02c9ddcc",
     "scripts/audit_goal_readiness.py": "4fe8f36c4ef9b8e3c492dcef894cabc7afe98b5396e1c4bd15bdcfef3da733d7",
     "scripts/baseer_ocr_extract.py": "056ca9cc33591db804639030a16d9635931b720d0d499b444ed6e7d0a653605a",
-    "scripts/benchmark_ocr.py": "b5ffb17845a7945b2a5c52e38bfabb6d82f3a8fbc8f2cdd5528843e09ad4deb4",
+    "scripts/benchmark_ocr.py": "48f4dfe4cdf2e8318abe968092cee7671989becaaea41e1b1498e2ac0261cccd",
     "scripts/benchmark_voices.py": "705bdfb6260fe90a4a68d9d2455953ea7221d282bbf0cc1cc4fa32cd5ed10205",
     "scripts/check_deployment_readiness.py": "c371706cf94f807354a1a08f274dc17b1c02d68347b98f70b177b4c14f73bf17",
     "scripts/check_research_sources.py": "49bc5a15cddf040f134d21e042d064d64fce2235f2ff1dd01f6b9c69cdf0c3e0",
@@ -97,8 +98,8 @@
     "scripts/verify_site.py": "7a09c02f0063f913ac76f0793dcf359684cb6d210c3c851e86934527b277295d",
     "scripts/verify_voice.py": "d8fb7e473e47060b2d2f957c5c230807a205e95b1469eef9c32b76d2bc8585b2",
     "scripts/verify_worker.py": "73329f87852ce805ab7144df6faaab4e081099f7ebc9a2e66e93735ee7fa82cc",
-    "static/app.js": "735d2ba288d8f96b7e99d4009d0ad5ef2db845562ea5defb5a6725b3c4dc6993",
-    "static/index.html": "0877f04c78afa4078c92fea23a93ff2f97851a8c3d17dd005e3c5a56b8508288",
+    "static/app.js": "5876611f1c9ce7f8e7df5c17fa57c7ba0632292b521b09ff6679a965f4d4a83d",
+    "static/index.html": "2fb1610ca4b5faa964e167c6d5d3d788352244b5ea53549ffe69962fcea58c5a",
     "static/styles.css": "a45485cf99eaae8a46e57437a736ce1ebad2528dbf219c5bc79f124ec3c47164"
   },
   "source": "ArabicTranslator",

app/main.py

@@ -119,6 +119,7 @@ OCR_ENGINE_CHOICES = {
     "paddleocr-vl",
     "surya",
     "tesseract",
+    "tesseract-fast",
     "auto",
     "best",
 }
@@ -283,10 +284,11 @@ CLOUD_VOICES = {
 LOCAL_VOICES = {
     "silma-local": {
         "id": "silma-local",
-        "label": "SILMA Arabic - Recommended voice",
+        "label": "1. SILMA Arabic - Most natural",
         "engine": "silma",
         "license": "Apache-2.0",
-        "recommendedFor": "Best quality/speed balance for free Arabic audio on this worker",
+        "recommendedFor": "Best Arabic accuracy/naturalness among voices generated from the winning OCR sample",
+        "rank": 1,
     },
     "habibi-msa": {
         "id": "habibi-msa",
@@ -303,17 +305,19 @@ LOCAL_VOICES = {
     },
     "espeak-ar-clear": {
         "id": "espeak-ar-clear",
-        "label": "Local Arabic Clear",
+        "label": "2. Local Arabic Clear - Fast fallback",
         "engine": "espeak-ng",
         "voice": "ar+f2",
         "license": "GPL-compatible open-source eSpeak NG",
-        "recommendedFor": "Fast fallback when neural voices are too slow or unavailable",
+        "recommendedFor": "Best generated fallback when SILMA is too slow or unavailable",
+        "rank": 2,
     },
     "espeak-ar": {
         "id": "espeak-ar",
-        "label": "Local Arabic",
+        "label": "3. Local Arabic - Standard fallback",
         "engine": "espeak-ng",
         "voice": "ar",
+        "rank": 3,
     },
     "espeak-ar-male": {
         "id": "espeak-ar-male",
@@ -323,6 +327,78 @@ LOCAL_VOICES = {
     },
 }
 
+OCR_BENCHMARK_RANKING = [
+    {
+        "rank": 1,
+        "id": "tesseract",
+        "label": "1. Tesseract Arabic - Best readable",
+        "extraction": "tesseract@2x-psm4",
+        "settings": "OCR_RENDER_ZOOM=2 TESSERACT_PSM=4",
+        "quality": "good",
+        "qualityScore": 11919.05,
+        "seconds": 37.30,
+        "arabicWords": 3120,
+        "note": "Most readable 5-page benchmark output; default for full-book runs.",
+    },
+    {
+        "rank": 2,
+        "id": "tesseract-fast",
+        "label": "2. Tesseract Arabic - Faster readable",
+        "extraction": "tesseract@1.5x-psm6",
+        "settings": "OCR_RENDER_ZOOM=1.5 TESSERACT_PSM=6",
+        "quality": "good",
+        "qualityScore": 11510.50,
+        "seconds": 28.88,
+        "arabicWords": 3284,
+        "note": "Runner-up readable setting; faster, but slightly lower text-quality score.",
+    },
+    {
+        "rank": 3,
+        "id": "paddleocr",
+        "label": "3. PaddleOCR Arabic - Faster fallback",
+        "extraction": "paddleocr",
+        "settings": "OCR_ENGINE=paddleocr",
+        "quality": "warning",
+        "qualityScore": 8105.80,
+        "seconds": 106.91,
+        "arabicWords": 2251,
+        "note": "Usable Arabic OCR fallback, but more fragmented on this book sample.",
+    },
+]
+
+VOICE_BENCHMARK_RANKING = [
+    {
+        "rank": 1,
+        "id": "silma-local",
+        "label": "1. SILMA Arabic - Most natural",
+        "engine": "silma",
+        "generated": True,
+        "elapsedSeconds": 277.34,
+        "sample": "outputs/ranked-voice-benchmark/silma-local.mp3",
+        "note": "Only generated neural Arabic voice in the benchmark; best starting point for actual Arabic naturalness.",
+    },
+    {
+        "rank": 2,
+        "id": "espeak-ar-clear",
+        "label": "2. Local Arabic Clear - Fast fallback",
+        "engine": "espeak-ng",
+        "generated": True,
+        "elapsedSeconds": 0.10,
+        "sample": "outputs/ranked-voice-benchmark/espeak-ar-clear.mp3",
+        "note": "Fastest clear fallback when the neural voice is too slow or unavailable.",
+    },
+    {
+        "rank": 3,
+        "id": "espeak-ar",
+        "label": "3. Local Arabic - Standard fallback",
+        "engine": "espeak-ng",
+        "generated": True,
+        "elapsedSeconds": 0.10,
+        "sample": "outputs/ranked-voice-benchmark/espeak-ar.mp3",
+        "note": "Standard eSpeak Arabic fallback; generated successfully but less natural than SILMA.",
+    },
+]
+
 
 def get_voice_catalog() -> dict[str, object]:
     return {
@@ -1251,7 +1327,7 @@ def get_engine_status() -> dict[str, object]:
             "easyocr": {"available": easyocr_ready, "label": "General Arabic OCR"},
             "paddleocr": {
                 "available": paddleocr_ready,
-                "label": "PaddleOCR Arabic - faster, less readable",
+                "label": "3. PaddleOCR Arabic - Faster fallback",
                 "trainedFor": "Arabic printed text",
                 "model": "arabic_PP-OCRv5_mobile_rec",
                 "recommendedFor": "Usable fallback, but the 5-page benchmark produced more fragmented text than Tesseract",
@@ -1313,11 +1389,18 @@ def get_engine_status() -> dict[str, object]:
             },
             "tesseract": {
                 "available": bool(tesseract_path),
-                "label": "Tesseract Arabic - Recommended readable",
+                "label": "1. Tesseract Arabic - Best readable",
                 "trainedFor": "Arabic printed text",
                 "recommendedFor": "Best readable output on the 5-page Arabic benchmark; uses OCR_RENDER_ZOOM=2 and TESSERACT_PSM=4 by default",
             },
+            "tesseractFast": {
+                "available": bool(tesseract_path),
+                "label": "2. Tesseract Arabic - Faster readable",
+                "trainedFor": "Arabic printed text",
+                "recommendedFor": "Second-best readable output on the 5-page benchmark; uses OCR_RENDER_ZOOM=1.5 and TESSERACT_PSM=6",
+            },
             "language": os.getenv("OCR_LANGUAGE", "ara"),
+            "ranking": OCR_BENCHMARK_RANKING,
         },
         "readyForArabic": bool(
             silma_installed or habibi_installed or supertonic_installed or (piper_path and piper_model_ready) or espeak_path
@@ -1338,6 +1421,7 @@ def get_engine_status() -> dict[str, object]:
             "audioStorage": "worker-local retained downloads",
             "benchmarkRule": "Run a representative 5-page Arabic sample before full-book audio.",
         },
+        "voiceRanking": VOICE_BENCHMARK_RANKING,
         "voices": get_voice_catalog(),
         "deployment": {
             "platform": "vercel" if IS_VERCEL else "local",
@@ -2957,6 +3041,8 @@ def ocr_pdf_text(pdf_path: Path, job: Job) -> str:
         engines = [ocr_pdf_text_with_surya, ocr_pdf_text_with_paddleocr, ocr_pdf_text_with_easyocr]
     elif requested == "tesseract":
         engines = []
+    elif requested == "tesseract-fast":
+        return ocr_pdf_text_with_tesseract(pdf_path, job, render_zoom=1.5, psm=6)
     else:
         engines = [ocr_pdf_text_with_paddleocr, ocr_pdf_text_with_easyocr]
 

docs/father-user-guide.md

@@ -8,7 +8,7 @@ This guide is for the person using the website, not for setup.
 2. Enter the access code.
 3. Choose the Arabic PDF.
 4. Leave **Voice** on the best Arabic voice unless someone tells you to change it.
-5. Leave **Text quality** on **Tesseract Arabic - Recommended readable** for a new scanned book.
+5. Leave **Text quality** on **1. Tesseract Arabic - Best readable** for a new scanned book.
 6. Leave **Pages** on **Quick test** first.
 7. Press **Create Audio**.
 8. Wait until the status says the audio is ready.
@@ -18,7 +18,7 @@ This guide is for the person using the website, not for setup.
 
 ## Which Text Quality To Choose
 
-Use **Tesseract Arabic - Recommended readable** first. It produced the most readable text in the 5-page Arabic OCR benchmark and is much faster than the comparison modes.
+Use **1. Tesseract Arabic - Best readable** first. It produced the most readable text in the 5-page Arabic OCR benchmark and is much faster than the comparison modes.
 
 Use **Arabic OCR comparison - slower** or **Maximum Arabic OCR - slower** on a short sample when the recommended option reads badly. They compare more OCR results and keep the cleanest text, but they can take much longer.
 
@@ -28,13 +28,15 @@ Use **KATIB Arabic OCR** when QARI is too slow or too heavy. It is also trained
 
 Use **Best scan test** only on a short sample. It is useful for deciding which OCR engine works best for one book, but it is too slow for most full books.
 
-Use **PaddleOCR Arabic - faster, less readable** only when Tesseract is unavailable or a short test sounds better for that book.
+Use **2. Tesseract Arabic - Faster readable** when speed matters and a short test still sounds correct.
+
+Use **3. PaddleOCR Arabic - Faster fallback** only when Tesseract is unavailable or a short test sounds better for that book.
 
 Use **Tesseract Arabic fallback** when the other options are broken or when a benchmark says Tesseract worked best for that book.
 
 ## If Something Fails
 
-If the app says the text quality is poor, do not make full-book audio yet. Try **Tesseract Arabic - Recommended readable**, then **Arabic OCR comparison - slower**, then **Best scan test** on a short sample.
+If the app says the text quality is poor, do not make full-book audio yet. Try **1. Tesseract Arabic - Best readable**, then **Arabic OCR comparison - slower**, then **Best scan test** on a short sample.
 
 If the first pages are title pages or blank pages, make a 5-page test PDF from better pages and test that before the full book.
 
@@ -42,7 +44,7 @@ If the website is running through a hosted worker, the job can keep running whil
 
 If the hosted website says downloadable cloud audio needs the worker, use **Read aloud** only as a temporary listening option. That browser voice can help you hear the text, but the normal **Download Audio** button comes back only after the worker is connected.
 
-If the audio sounds robotic but the text is correct, try another voice. **SILMA Arabic - Recommended voice** is the preferred free worker voice when it is available.
+If the audio sounds robotic but the text is correct, try another voice. **1. SILMA Arabic - Most natural** is the preferred free worker voice when it is available. **2. Local Arabic Clear - Fast fallback** is the next choice when SILMA is too slow.
 
 ## Best Routine For A New Book
 

docs/ocr-and-voice-ranking.md

@@ -0,0 +1,35 @@
+# Arabic OCR and Voice Ranking
+
+Last run: June 8, 2026.
+
+PDF benchmark file: `test_pdfs/arabic-reader-5-page-test.pdf`
+
+Voice sample text: `outputs/ranked-voice-ocr-sample.txt`
+
+The voice samples were generated from the winning OCR output, so every voice read the same cleaned Arabic text.
+
+## OCR Ranking
+
+| Rank | Website option | Extraction | Quality | Score | Time | Arabic words | Result |
+| ---: | --- | --- | --- | ---: | ---: | ---: | --- |
+| 1 | `1. Tesseract Arabic - Best readable` | `tesseract@2x-psm4` | good | 11919.05 | 37.30s | 3120 | Best readable output; use by default. |
+| 2 | `2. Tesseract Arabic - Faster readable` | `tesseract@1.5x-psm6` | good | 11510.50 | 28.88s | 3284 | Runner-up; use when it sounds correct and speed matters. |
+| 3 | `3. PaddleOCR Arabic - Faster fallback` | `paddleocr` | warning | 8105.80 | 106.91s | 2251 | Usable fallback, but more fragmented on this book sample. |
+
+`EasyOCR` and `Auto fallback` both fell through to the same PaddleOCR output on this setup, so they are not ranked above the direct PaddleOCR option.
+
+## Voice Ranking
+
+Generated files:
+
+| Rank | Website option | Engine | Generated | Time | File | Result |
+| ---: | --- | --- | --- | ---: | --- | --- |
+| 1 | `1. SILMA Arabic - Most natural` | SILMA | yes | 277.34s | `outputs/ranked-voice-benchmark/silma-local.mp3` | Best Arabic-naturalness starting point because it is the only generated neural Arabic voice in this run. |
+| 2 | `2. Local Arabic Clear - Fast fallback` | eSpeak NG | yes | 0.10s | `outputs/ranked-voice-benchmark/espeak-ar-clear.mp3` | Best fast fallback when SILMA is too slow or unavailable. |
+| 3 | `3. Local Arabic - Standard fallback` | eSpeak NG | yes | 0.10s | `outputs/ranked-voice-benchmark/espeak-ar.mp3` | Standard fallback; generated successfully but is less natural than SILMA. |
+
+`Habibi Arabic MSA` and `Supertonic Arabic CPU` were listed but did not generate audio in this local run, so they are unranked until installed and tested on the same OCR sample.
+
+## Recommendation
+
+Use `1. Tesseract Arabic - Best readable` with `1. SILMA Arabic - Most natural` for the first full-book attempt. If SILMA is too slow, keep the OCR setting and switch the voice to `2. Local Arabic Clear - Fast fallback`.

docs/ocr-readability-benchmark.md

@@ -18,6 +18,12 @@ TESSERACT_PSM=4
 
 This setting produced the most readable 5-page output while staying practical for full-book jobs.
 
+Top 3 tested OCR settings:
+
+1. `Tesseract Arabic - Best readable`: `OCR_ENGINE=tesseract OCR_RENDER_ZOOM=2 TESSERACT_PSM=4`
+2. `Tesseract Arabic - Faster readable`: `OCR_ENGINE=tesseract-fast OCR_RENDER_ZOOM=1.5 TESSERACT_PSM=6`
+3. `PaddleOCR Arabic - Faster fallback`: `OCR_ENGINE=paddleocr`
+
 | OCR setting | Pages | Seconds | Quality | Score | Arabic words | Fragment line ratio | Extraction |
 | --- | ---: | ---: | --- | ---: | ---: | ---: | --- |
 | Tesseract 2x PSM 4 | 5 | 37.30 | good | 11919.05 | 3120 | 0.0433 | `tesseract@2x-psm4` |
@@ -39,4 +45,4 @@ The slower comparison modes were tested on the 1-page sample because the full 5-
 
 PaddleOCR is available and works, but on this book sample it returned many low-information lines and more fragmented Arabic text. It remains a fallback, not the recommendation.
 
-The live/default website setting should therefore be `Tesseract Arabic - Recommended readable`.
+The live/default website setting should therefore be `1. Tesseract Arabic - Best readable`.

docs/production-worker-architecture.md

@@ -93,7 +93,7 @@ The worker bundle also includes setup scripts for optional heavy paths, but they
 
 `Dockerfile.worker` exposes `INSTALL_QARI_OCR=1`, `INSTALL_KATIB_OCR=1`, `INSTALL_ARABIC_QWEN_OCR=1`, `INSTALL_BASEER_OCR=1`, `INSTALL_PADDLEOCR_VL=1`, and `INSTALL_SUPERTONIC=1` build args so stronger workers can install QARI-OCR, KATIB, Arabic-Qwen, Baseer, PaddleOCR-VL, and the optional Supertonic CPU voice without editing the Dockerfile. This keeps the free CPU image practical while making the higher-quality free OCR paths and fast voice comparison path deployable.
 
-QARI-OCR 0.4 is the strongest Arabic-native OCR upgrade to test for a stronger worker. It is a 4B VLM fine-tuned for Islamic books and Arabic manuscripts, so keep it out of the default free CPU family-site worker unless a short Arabic-book benchmark proves it improves the actual pages and the worker has enough RAM/GPU. KATIB 0.8B and Arabic-Qwen3.5-OCR-v4 are the smaller Arabic-trained OCR upgrades to try when QARI is too heavy. If the worker is too small for QARI, set `QARI_OCR_MODEL=NAMAA-Space/Qari-OCR-v0.3-VL-2B-Instruct` to test the lighter older QARI path. PaddleOCR-VL-1.6 remains the main general document-parser upgrade to watch. In the website, start with `Tesseract Arabic - Recommended readable`; use `Arabic OCR comparison - slower` or `Maximum Arabic OCR - slower` only for short tests or difficult pages.
+QARI-OCR 0.4 is the strongest Arabic-native OCR upgrade to test for a stronger worker. It is a 4B VLM fine-tuned for Islamic books and Arabic manuscripts, so keep it out of the default free CPU family-site worker unless a short Arabic-book benchmark proves it improves the actual pages and the worker has enough RAM/GPU. KATIB 0.8B and Arabic-Qwen3.5-OCR-v4 are the smaller Arabic-trained OCR upgrades to try when QARI is too heavy. If the worker is too small for QARI, set `QARI_OCR_MODEL=NAMAA-Space/Qari-OCR-v0.3-VL-2B-Instruct` to test the lighter older QARI path. PaddleOCR-VL-1.6 remains the main general document-parser upgrade to watch. In the website, start with `1. Tesseract Arabic - Best readable`; use `Arabic OCR comparison - slower` or `Maximum Arabic OCR - slower` only for short tests or difficult pages.
 
 The repo includes optional KATIB, Arabic-Qwen, QARI-OCR, and PaddleOCR-VL sidecar scripts for this evaluation path:
 

scripts/benchmark_ocr.py

@@ -132,6 +132,16 @@ def recommendation_for_extraction(extraction: str | None) -> dict[str, Any] | No
             env["OCR_RENDER_ZOOM"] = zoom
     if psm and engine == "tesseract":
         env["TESSERACT_PSM"] = psm
+    if engine == "tesseract-fast":
+        env["OCR_ENGINE"] = engine
+        env["OCR_RENDER_ZOOM"] = zoom or "1.5"
+        env["TESSERACT_PSM"] = psm or "6"
+        notes.append("Use this runner-up setting when speed matters and its sample text still sounds correct.")
+        return {
+            "summary": "For the full book, use OCR_ENGINE=tesseract-fast OCR_RENDER_ZOOM=1.5 TESSERACT_PSM=6.",
+            "env": env,
+            "notes": notes,
+        }
     if engine == "tesseract":
         notes.append("Confirm Tesseract Arabic data is installed before the full run.")
     elif engine == "easyocr":
@@ -169,7 +179,11 @@ def benchmark_engine(pdf_path: Path, engine: str) -> dict[str, Any]:
     job = main.Job(id="dry-run", filename=pdf_path.name, ocr_engine=engine)
     started = time.perf_counter()
     try:
-        text = main.extract_pdf_text(pdf_path, job)
+        if engine == "tesseract-fast":
+            text = main.ocr_pdf_text_with_tesseract(pdf_path, job, render_zoom=1.5, psm=6)
+            job.ocr_engine = engine
+        else:
+            text = main.extract_pdf_text(pdf_path, job)
         elapsed = round(time.perf_counter() - started, 2)
         result = {
             "engine": engine,
@@ -179,7 +193,18 @@ def benchmark_engine(pdf_path: Path, engine: str) -> dict[str, Any]:
             "extraction": job.extraction,
             **text_metrics(text),
         }
-        result["recommendation"] = recommendation_for_extraction(job.extraction)
+        if engine == "tesseract-fast":
+            result["recommendation"] = {
+                "summary": "For the full book, use OCR_ENGINE=tesseract-fast.",
+                "env": {
+                    "OCR_ENGINE": "tesseract-fast",
+                    "OCR_RENDER_ZOOM": "1.5",
+                    "TESSERACT_PSM": "6",
+                },
+                "notes": ["Use this runner-up setting when speed matters and its sample text still sounds correct."],
+            }
+        else:
+            result["recommendation"] = recommendation_for_extraction(job.extraction)
         return result
     except Exception as exc:
         elapsed = round(time.perf_counter() - started, 2)
@@ -244,6 +269,7 @@ def main_cli() -> None:
             "paddleocr-vl",
             "surya",
             "tesseract",
+            "tesseract-fast",
             "auto",
             "best",
         ],

static/app.js

@@ -64,9 +64,9 @@ const defaultVoiceCatalog = {
     { id: "silma-tts", label: "SILMA Arabic" },
   ],
   local: [
-    { id: "silma-local", label: "SILMA Arabic - Recommended voice" },
-    { id: "espeak-ar-clear", label: "Local Arabic Clear" },
-    { id: "espeak-ar", label: "Local Arabic" },
+    { id: "silma-local", label: "1. SILMA Arabic - Most natural" },
+    { id: "espeak-ar-clear", label: "2. Local Arabic Clear - Fast fallback" },
+    { id: "espeak-ar", label: "3. Local Arabic - Standard fallback" },
     { id: "espeak-ar-male", label: "Local Arabic Low" },
   ],
 };
@@ -83,6 +83,9 @@ let browserSpeechText = "";
 let browserSpeechSourceName = "";
 
 const ocrModeLabels = {
+  tesseract: "1. Tesseract Arabic - Best readable",
+  "tesseract-fast": "2. Tesseract Arabic - Faster readable",
+  paddleocr: "3. PaddleOCR Arabic - Faster fallback",
   "arabic-max": "Maximum Arabic OCR - slower",
   arabic: "Arabic OCR comparison - slower",
   "qari-ocr": "QARI Arabic books (best)",
@@ -91,12 +94,10 @@ const ocrModeLabels = {
   "arabic-qwen-ocr": "Arabic-Qwen OCR",
   "arabic-glm-ocr": "Arabic-GLM OCR v2",
   "baseer-ocr": "Baseer Arabic OCR",
-  paddleocr: "PaddleOCR Arabic - faster, less readable",
   "paddleocr-vl": "PaddleOCR-VL heavy",
   best: "Best scan test",
   surya: "Surya heavy OCR",
   easyocr: "General Arabic OCR",
-  tesseract: "Tesseract Arabic - Recommended readable",
   auto: "Auto fallback",
 };
 
@@ -336,7 +337,9 @@ async function loadHealth() {
           : engines.ocr?.preferred === "surya"
             ? "Surya heavy OCR is ready"
           : engines.ocr?.preferred === "tesseract"
-            ? "Recommended readable Tesseract Arabic OCR is ready"
+            ? "Rank 1 readable Tesseract Arabic OCR is ready"
+          : engines.ocr?.preferred === "tesseract-fast"
+            ? "Rank 2 faster Tesseract Arabic OCR is ready"
           : engines.ocr?.preferred === "best"
             ? "Best Arabic OCR test mode is ready"
             : engines.ocr?.preferred
@@ -1415,10 +1418,13 @@ function describeOcrMode() {
     engineNotice.textContent = "Best scan test selected. Use this on a short sample, then run the winning engine for the full book.";
     engineNotice.classList.remove("warning");
   } else if (ocrModeSelect.value === "paddleocr") {
-    engineNotice.textContent = "PaddleOCR Arabic selected. It works, but the 5-page benchmark produced more fragmented text than Tesseract.";
+    engineNotice.textContent = "Rank 3 PaddleOCR Arabic selected. It works, but the 5-page benchmark produced more fragmented text than Tesseract.";
     engineNotice.classList.remove("warning");
   } else if (ocrModeSelect.value === "tesseract") {
-    engineNotice.textContent = "Tesseract Arabic selected. This is the recommended readable option from the 5-page OCR benchmark.";
+    engineNotice.textContent = "Rank 1 Tesseract Arabic selected. This produced the best readable text in the 5-page OCR benchmark.";
+    engineNotice.classList.remove("warning");
+  } else if (ocrModeSelect.value === "tesseract-fast") {
+    engineNotice.textContent = "Rank 2 Tesseract Arabic selected. This was the faster readable runner-up in the 5-page OCR benchmark.";
     engineNotice.classList.remove("warning");
   } else if (ocrModeSelect.value === "paddleocr-vl") {
     engineNotice.textContent = "PaddleOCR-VL selected. Use this only on a short sample or strong worker; it is much heavier than normal Arabic OCR.";
@@ -1649,8 +1655,8 @@ function showQualityHint(quality) {
   }
   const reasons = quality.reasons?.length ? ` ${quality.reasons.join("; ")}.` : "";
   const action = quality.quality === "poor"
-    ? "Try Tesseract Arabic - Recommended readable, Best scan test, or another OCR mode before creating audio."
-    : "Listen to a short sample before running the full book. If it sounds wrong, try Tesseract Arabic - Recommended readable, Best scan test, or another OCR mode.";
+    ? "Try 1. Tesseract Arabic - Best readable, Best scan test, or another OCR mode before creating audio."
+    : "Listen to a short sample before running the full book. If it sounds wrong, try 1. Tesseract Arabic - Best readable, Best scan test, or another OCR mode.";
   qualityHint.textContent = `Text needs checking.${reasons} ${action}`;
   qualityHint.classList.remove("hidden");
   qualityHint.classList.toggle("poor", quality.quality === "poor");

static/index.html

@@ -67,7 +67,9 @@
           <div class="field-group">
             <label for="ocrModeSelect">Text quality</label>
             <select id="ocrModeSelect" name="ocrMode">
-              <option value="tesseract">Tesseract Arabic - Recommended readable</option>
+              <option value="tesseract">1. Tesseract Arabic - Best readable</option>
+              <option value="tesseract-fast">2. Tesseract Arabic - Faster readable</option>
+              <option value="paddleocr">3. PaddleOCR Arabic - Faster fallback</option>
               <option value="arabic">Arabic OCR comparison - slower</option>
               <option value="arabic-max">Maximum Arabic OCR - slower</option>
               <option value="qari-ocr">QARI Arabic books (best)</option>
@@ -77,7 +79,6 @@
               <option value="arabic-glm-ocr">Arabic-GLM OCR v2</option>
               <option value="baseer-ocr">Baseer Arabic OCR</option>
               <option value="best">Best scan test</option>
-              <option value="paddleocr">PaddleOCR Arabic - faster, less readable</option>
               <option value="paddleocr-vl">PaddleOCR-VL heavy</option>
               <option value="surya">Surya heavy OCR</option>
               <option value="easyocr">General Arabic OCR</option>