{"result":"The image contains a black background with white text. The text appears to be a transcript or log file, likely from a data analysis or software process, as indicated by the presence of timestamps, variable names (X1, X2, X3, Z, Y), statistical terms (beta, regression, scatter plot, p-value, T-Statistik, Koefisien), and phrases like \"ANALISIS REGRESI PATH (PLS-SEM)\", \"Responden\", and \"Uji Hipotesis\".\n\nThe content discusses a path analysis with variables X1, X2, X3, Z, and Y, and their relationships. It mentions scatter plots, regression lines (colored purple, green, and blue), and path diagrams. The analysis seems to involve evaluating direct and indirect effects, with beta coefficients, p-values, and T-Statistics reported for various paths. There are also references to tables summarizing these findings, such as \"Ringkasan Lengkap Hasil Uji Hipotesis\".\n\nThe timestamps and the format of the output suggest it's a log of operations or results displayed sequentially. The text also indicates that the output is divided into sections, with \"Tampilan terbagi dua\" (display divided in two) appearing multiple times, referring to left (diagram/scatter plot) and right (diagram/scatter plot) panels.","media_info":{"size_mb":0.61,"duration_sec":null,"approx_tokens":300},"style":null}
curl --location 'https://zylalabs.com/api/13135/multimodal+transcript+api/26649/multimodal+transcript+api' \
--header 'Content-Type: application/json' \
--form 'image=@"FILE_PATH"'
注册后,每个开发者都会被分配一个个人 API 访问密钥,这是一个唯一的字母和数字组合,用于访问我们的 API 端点。要使用 Multimodal Transcript API 进行身份验证,只需在 Authorization 标头中包含您的 bearer token。
| 标头 | 描述 |
|---|---|
授权
|
必需
应为 Bearer access_key. 订阅后,请查看上方的"您的 API 访问密钥"。
|
无长期承诺。随时升级、降级或取消。 免费试用包括最多 50 个请求。
(年度计费可节省 2 个月 🎉)
领先企业的信赖之选
Multimodal Transkrip turns any image, audio clip, or video into rich, structured insight through a single unified API. Instead of juggling separate services for vision, speech, and video understanding, send us the file and get back a detailed, human-readable analysis in seconds.
Key capabilities:
- Image analysis: object/scene detection, text recognition, contextual description.
- Audio analysis: transcription plus tone and notable-sound summarization.
- Video analysis: scene-by-scene description, either as a single flowing summary or as timestamped SRT-style subtitle output — useful even for silent video.
- Multimodal endpoint: combine image, audio, and video in one request for a unified analysis.
- Custom narration style: request the result be delivered in a specific tone or persona (e.g. news reporter, sports commentator, gentle narrator) separate from what gets analyzed.
Billing scales fairly with actual media size and duration, so a small image and a 30-minute video aren't charged the same. Built for developers doing content moderation, accessibility captioning, video indexing/search, or automated media cataloging.
Each endpoint returns structured insights based on the analyzed media. For images, it provides object detection, text recognition, and contextual descriptions. Audio analysis yields transcriptions and tone summaries, while video analysis offers scene descriptions and timestamped subtitles.
Key fields in the response data include "result," which contains detailed descriptions of detected objects, text, audio transcriptions, and scene summaries. Each field provides insights relevant to the specific media type analyzed.
The returned data is structured in JSON format. It typically includes a "result" field containing a detailed narrative of the analysis, which may include lists of detected items, transcriptions, and contextual information relevant to the media.
Each endpoint provides information such as object and scene descriptions for images, transcriptions and tone analysis for audio, and detailed scene-by-scene descriptions or subtitles for videos, allowing for comprehensive media understanding.
Users can customize requests by specifying parameters such as the desired narration style or tone for the output. This allows for tailored responses, such as requesting a formal report or a casual commentary style.
The response data is organized in a hierarchical JSON structure, with the main "result" field containing detailed insights. This structure allows users to easily parse and utilize specific information relevant to their needs.
Typical use cases include content moderation, accessibility captioning, video indexing and search, and automated media cataloging. The API's capabilities support diverse applications across various industries.
Data accuracy is maintained through advanced AI algorithms that analyze media content. Continuous updates and improvements to the models ensure high-quality outputs, while user feedback may also contribute to refining the analysis processes.