Rejaviy imtiyozlar

Addon T

Qo'shimcha xususiyat foydalanuvchilari uchun ixtiyoriy reja

Addon T obuna bo'yicha qo'llanma

Obuna bo'yicha qo'llanma

Xususiyat bo'yicha qo'shimcha to'lov solishtirish (-25%)

Review successful usage only. Math similar-question generation applies academy contract prices first; Questi uses actual user plan/generation prices.

{billableCount} billable items / {totalCount} total items

Billing by feature

Billing Invoice

View Book Chat web usage trends by day/week/month/custom range.

Book Chat Trend

View library/Book Chat OCR analysis usage trends by day/week/month/custom range.

Book OCR Analysis Trend

Operator boshqaruv paneli va foydalanuvchi boshqaruvi

Administrator konsoli

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha foydalanish tendentsiyasini koʻring.

Qiyinchilik tendentsiyasi

Ishlab chiqarish/koʻrib chiqish tendentsiyalari va asosiy operatsion koʻrsatkichlarni bir joyda koʻring.

Shu kabi muammoli avlod tendentsiyalarini tahlil qiling

Kontentni tahrirlash

Qo'llash bosilganda o'zgarishlar sahifaga qo'llanadi.

Formulani tahrirlash

Avlod

Mening DB ro‘yxatdan o‘tish tendensiyasini tahlil qiling

Mening DB

Ko'rib chiqishdan foydalanish tendentsiyalarini tahlil qiling

Ko‘rib chiqish

Foydalanuvchilar roʻyxati va funksiya cheklovlarini boshqaring

Foydalanuvchi boshqaruvlari

Persisted activity in the selected period, using the measure shown on each card. Generation, editing and grading requests count separately.

Administrator boshqaruv paneli

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha repetitorning foydalanish tendentsiyasini 1:1 koʻrinishida koʻring.

1:1 Repetitor trendi

Rasmlardan shablon yaratish so‘rovlarining bajarilish natijalarini, shuningdek, haqiqiy API chaqiruvlari, tokenlar va xarajatlarni ko‘ring. Har bir so‘rov ko‘pi bilan 5 sahifani o‘z ichiga oladi.

Rasm→shablon foydalanish statistikasi

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha hujjat konvertatsiyasidan foydalanish tendentsiyasini koʻring.

Hujjatlarni o'zgartirish tendentsiyasi

Review analysis times, requesters, indexing attempts and errors per document. Tokens and costs not reported by File Search remain unavailable.

Book OCR analysis

Review segmentation, generation and modification tokens, calls, agents, costs and errors per solution history.

Handwritten solutions

Historical usage statistics. Past token usage and costs are not backfilled.

User downloads

Kun/hafta/oy/moslashtirilgan diapazonda inshoni tuzatishdan foydalanish tendentsiyasini ko'ring.

Inshoni tuzatish tendentsiyasi

View exam similarity analysis usage trends by day/week/month/custom range.

Exam Similarity Analysis Trend

Kun/hafta/oy/moslashtirilgan diapazonda imtihon tahlilidan foydalanish tendentsiyasini ko'ring.

Imtihonlarni tahlil qilish tendentsiyasi

Review and collect school and mock exam inputs from document conversion, exam analysis, and similarity analysis.

Review and collect exam material

Subject

Exam material collection

Kunlar bo‘yicha natijalar

Har kuni bo‘limlar bo‘yicha muvaffaqiyat va nosozliklarni ko‘ring, nosoz akkauntlar va sabablarni tekshiring, tuzatgandan so‘ng xabarnoma yuboring.

{count} akkauntga xat yuboriladi.

Hal qilingani haqida xabar

Asl xato saqlanmagan. Ko‘rsatilgan izoh va ushbu vazifaning ishchi jurnallarini tekshiring.

Nosozliklar konsoli

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha topilgan savol (manba) foydalanish tendentsiyasini koʻring.

Savol tendentsiyasini toping

Usage and request counts for the selected period.

AI mathematics problem generation

View per-run records and existing usage statistics in one place.

View historical usage and request counts for this subject. Legacy and V2 are different views, not a split by creation date.

View English similar problem generation usage by hour, day, week, month, or a custom period.

English Similar Problems Usage Trend

View Korean similar problem generation usage by hour, day, week, month, or a custom period.

Korean Similar Problems Usage Trend

View math similar problem generation usage by hour, day, week, month, or a custom period.

Math Similar Problems Usage Trend

Inspect each history's actor, LLM usage, cost, and errors from stored operational metadata.

Individual call records load only when details are expanded.

LLM call records

Run details

Errors across all matching runs, ordered by frequency.

Most frequent errors

Runs with stored observability metadata / all runs

Operations overview

Run history

Problem generation runs

Science problem generation

View STEAM problem generation usage trends by day/week/month/custom range.

STEAM Problem Trend

Kun/hafta/oy/moslashtirilgan davr boʻyicha shunga oʻxshash muammo yaratish tendentsiyalarini koʻring.

Avloddan foydalanish

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha qoʻlda yozilgan yechimdan foydalanish tendentsiyasini koʻring.

Qo'lda yozilgan yechim tendentsiyasi

View Labeling API usage trends by day/week/month/custom range.

Labeling API Trend

Kun/hafta/oy/moslashtirilgan diapazonda matematik manipulyativ foydalanish tendentsiyasini ko'ring.

Matematik vositasi trendi

Manbalarni boshqaring, konveyerni ishga tushiring, yangilikni qo‘lda joylang.

Sarlavha va havolani kiriting — AI qisqacha mazmun, muhimlik va turkumni tayyorlaydi. Tekshirib joylang.

Joylanadigan sarlavha

Asl sarlavha

E’lon sarlavhasini o‘zgartirmasdan joylashtiring

Qo‘lda kiritish

Bir marta: yig‘ish → takrorni olib tashlash → AI saralash → rasmiy matn bilan solishtirish → joylash → bildirishnoma.

Yig‘ish konveyeri

Manba

Faqat lentasi bor va robots.txt ruxsat bergan manbalarni qo‘shing.

Manbalar

Ta’lim yangiliklarini boshqarish

Kun/hafta/oy/maxsus oraliq bo‘yicha Mening DB ro‘yxatdan o‘tish tendensiyasini ko‘ring.

Mening DB tendensiyasi

Legacy shows historical exam analysis usage. V2 combines measured tokens, cost, agents, and errors from generation, modifications, and retries for each history. Earlier calls are not backfilled.

Kun/hafta/oy/moslashtirilgan davr bo'yicha foydalanish tendentsiyalarini ko'rib chiqing.

Foydalanishni ko'rib chiqing

Kun/hafta/oy/moslashtirilgan diapazon bo'yicha ob'ektni ko'rib chiqish va yorliqlardan foydalanish tendentsiyasini ko'ring.

Ko‘rib chiqish/yorliq trendi

Combine segmentation, child video generation, and retry tokens, LLM/TTS calls, API cost, and errors per original request. API cost covers metered calls; rendering and storage costs are excluded.

Legacy shows historical video generation requests and completed videos. V2 shows newly measured API usage per history. Earlier calls are not backfilled.

Completed videos and original requests grouped by request time. Child histories do not duplicate the request count.

Video generation

Nom bo'yicha

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha sintaktik tahlildan foydalanish tendentsiyasini koʻring.

Sintaktik tahlil tendentsiyasi

Admin-only Claude document conversion process

Read the whole page, divide questions, exclude doodles and redraw figures as SVG.

Document OCR

Convert the generated SVG into a GeoGebra file.

SVG to GeoGebra (optional)

Create a solution and answer for each question, plus a solution SVG if needed.

Generate solutions (optional)

Tag curriculum, format and concepts; save question number and source in the private bank.

Tag and save each question

Document conversion process Claude 261020

The run log shows LLM calls and tokens by History ID. Its cost total includes only priced models; currently unpriced calls such as Gemini Nano Banana 2.1 image generation need a separate estimate.

Cost per run

Document conversion flow and stage models as of October 20, 2026

OCR and problem tagging may use GPT-5.5 on the final retry after Gemini failures. Figure detection or OCR may use GPT-6.1 Sol on the final retry after a Gemini 503 error. Retry calls also add cost.

Retries and fallback models

Standard conversion connects figure references by detection_id and marks problem boundaries with problem-break. Solution mode maps figures to problem crops by geometry.

Linking figures to problems

These calls run when enabled and a target figure exists. GeoGebra generation also requires USE_AUTOMATIC_GGB_GENERATION on the server.

Optional figure processing

Problems are split into crops, tagged, OCRed, and solved individually. Successful problems are stored separately in problem_banks.

Generated solution mode

Document type

Identify candidates for SVG and GeoGebra conversion.

Figure classification

Find figure regions and detection IDs in each source image.

Figure detection

Refine OCR with figure crops when figures exist.

Figure-aware OCR refinement

Convert SVG when both the user option and server setting are enabled.

GeoGebra conversion

When enabled, remove handwriting or doodles from detected figures.

Handwriting removal

Extract text, formulas, problem boundaries, and figure references.

Initial OCR

Create per-problem crops, including problems spanning pages.

Problem segmentation

Per-problem solution

Extract problem text and tags from each crop.

Per-problem tagging and OCR

Reconstruct selected figures as vector SVG.

SVG reconstruction

Each PDF page or uploaded image is processed as a source. Results are stored by source in typing_history; this path does not create one database record per problem.

Standard conversion

문서변환 프로세스 261020

Inspect each conversion’s operator, results, tokens, API cost, and errors.

View existing document conversion usage and request counts over time. Legacy and V2 are different views.

Bir yoki bir nechta foydalanuvchi bo‘yicha yaratish/ko‘rib chiqish/Mening DB tendensiyalarini solishtiring.

Foydalanuvchi statistikasi

Kredit o‘zgarishlari

Sahifalangan foydalanuvchilarni koʻrib chiqing, kreditlar bering va har bir foydalanuvchi uchun funksiya cheklovlarini boshqaring.

Maxsus xususiyat qo'shing

Ruxsat etilgan qo'shimcha funksiyalar

Foydalanuvchi kreditlarini ayirish

Foydalanuvchi kreditlarini berish

Foydalanuvchi xususiyatlari cheklovlarini o'zgartirish

Foydalanuvchi a'zoligini o'zgartirish

Obuna tarixi

Barcha foydalanuvchilar

Foydalanuvchilar

Foydalanuvchining foydalanishi haqida xulosa

View lecture video analysis usage trends by day/week/month/custom range.

Lecture Video Analysis Trend

View English vocabulary builder usage trends by day/week/month/custom range.

Vocab Builder Trend

Kun/hafta/oy/moslashtirilgan diapazon boʻyicha taʼlim webtoonidan foydalanish tendentsiyasini koʻring.

Edu Webtoon trendi

Gemini 3.8 Flash plans SVG and GGB corrections; gpt-image-2.5-flare draws the corrected figures. GGB corrections reference the previous SVG correction plan and image. Figures are stored as images, without individually editable points or lines.

Option 2-1 · Gemini + GPT Image 2.5 Flare

Highest difficulty design, independent review and accurate figures · isolated prompts

Advanced: import killer configuration

Killer option

Keep the original image and OCR; skip original SVG reconstruction and cached SVG reuse. Generate and validate the transformed figure with the same option 1-1 settings.

Process argon1014

Previous record: QuestiGeo 1014_1023

Classify the selected crop as geometry, algebraic, physics, or other and cache the result. Classification no longer shows a warning.

Classify before upload

Save each figure crop. Show the figure count multiplied by 500 credits and start processing only after the user agrees to the deduction.

Crop each figure and confirm 500 credits

Convert SVG1 to GeoGebra XML and GGB1, and show GGB1 as soon as it is ready.

SVG1 to GGB1

Load GGB1 and inspect its quality. Create GGB2 and replace the active GGB only when the corrected result is selected; retain GGB1 in history.