GET /{pullzone_name}/{video_external_id}-ai.{lang}.json

Returns the AI resources generated for a video in one language: eBook (abstract), Mind Map (mindmap) and Quiz (questions). This is the file an integrator reads to build their own interface — the Panda player consumes exactly these files.

In the AI Package v2 there is one file per language, identified by the .{lang} suffix in the name. Each file carries schema_version: 2 and only the keys of the resources that exist in that language. A video with an eBook in English and a Quiz in Portuguese has two files, each with its own block.

📘

Authorization

None. The file is served by the configuration CDN and does not require an authentication header. Do not send your API key in this call.

Naming reference

The API uses technical names; the Panda interface uses commercial names. In fields and enum values, always the technical ones.

HTTP Method & Path

GET /{pullzone_name}/{video_external_id}-ai.{lang}.json

Base URL: https://config.tv.pandavideo.com.br

Path Parameters

How to find out which files exist

⚠️

Do not build the path by trial and error. The correct entry point is the video manifest: config.ai.resources[] lists the type and language of every existing resource. Requesting a language that does not exist returns 404.

1. Call GET /videos/{video_id} and read config.ai.resources[]:

{
  "ai": {
    "abstract": true,
    "mindmap": true,
    "questions": true,
    "pdf_generated": true,
    "resources": [
      { "type": "abstract",  "srclang": "en" },
      { "type": "abstract",  "srclang": "pt-BR", "is_legacy": true },
      { "type": "questions", "srclang": "pt-BR", "is_legacy": true },
      { "type": "mindmap",   "srclang": "pt-BR", "is_legacy": true }
    ]
  }
}

2. For every entry without is_legacy, build the path using its srclang in place of {lang}. In the example above, only the English eBook is v2:

GET /{pullzone_name}/{video_external_id}-ai.en.json

3. Entries flagged is_legacy: true are read at the path without a language suffix — see the Legacy section at the end of this page. Entries flagged is_created: true were created by hand instead of generated by AI, and are read at the normal language-suffixed path.

Response

Success Response (200) — file with an eBook

{
  "schema_version": 2,
  "lang": "en",
  "abstract": "# Chapter 1\n\nThe first chapter in markdown...",
  "abstract_html": "<section class=\"panda-ebook\">...</section>",
  "abstract_summary": "A short summary of the whole eBook.",
  "abstract_cover": {
    "kicker": "Biochemistry",
    "title": "Vitamins",
    "subtitle": "Functions, sources and deficiencies",
    "edition_label": "Edition 1"
  },
  "abstract_config": {
    "layout": "modern",
    "title_font": "Rubik",
    "content_font": "Rubik",
    "accent_color": "#3F55EE",
    "secondary_color": "#E12AFB",
    "format": "portrait",
    "header_footer": true,
    "logo_key": null
  },
  "images": [
    {
      "storage_path": "USER_ID/ai_abstract_image/{uuid}.jpeg",
      "cdn_base": "https://images-cdn.pandavideo.com",
      "caption": "Molecular structure of vitamin C"
    }
  ]
}

Success Response (200) — file with a Quiz

The same video, another language, another resource. Note that there is no eBook key here, and that questions_config is a sibling of questions:

{
  "schema_version": 2,
  "lang": "pt-BR",
  "questions": [
    {
      "id": "{question_id}",
      "enunciate": "Which vitamin is synthesised through sun exposure?",
      "options": [
        { "text": "Vitamin A", "correct": false },
        { "text": "Vitamin D", "correct": true },
        { "text": "Vitamin K", "correct": false }
      ]
    }
  ],
  "questions_config": {
    "num_questions": 5,
    "num_answers": 3,
    "allow_skip": false,
    "show_explanation": true,
    "show_results": true,
    "collect_lead": false
  }
}

Response Schema

The fields of abstract_config and questions_config are detailed on Generate an AI Resource.

⚠️

Important

• The file carries only the keys of the resource that exists in that language — do not treat a missing field as an error.
• questions_config is a sibling of questions. Do not look for it inside the array.
• pdf_url does not exist in this file. The PDF URL is returned only by GET /videos/{video_id}, inside config.ai.resources[]. This is deliberate: in the public file it would be a direct download link, bypassing the per-resource download permission.
• abstract_html is the canonical source of the eBook. If you render the abstract markdown yourself, the result may differ from the PDF.

Error Responses

• 404: no resource exists for this video in this language, or the language suffix is wrong (e.g. pt instead of pt-BR). Check config.ai.resources[] instead of guessing the path.
• 403: invalid pullzone or video_external_id.

Legacy (v1) — -ai.json

The old path keeps working and is still written. Content generated with the previous model appears in the manifest flagged is_legacy: true and is read without a language suffix:

GET /{pullzone_name}/{video_external_id}-ai.json

🚧

Migration

Nothing already published became invalid — it became incomplete. Anyone reading only -ai.json will not find content generated with v2. Move to entering through the manifest (config.ai.resources[]) and handling both paths.

Example Usage

cURL

# v2 — eBook in English
curl -X GET "https://config.tv.pandavideo.com.br/vz-163a34ba-25b/{video_external_id}-ai.en.json"

# v2 — Quiz in Portuguese
curl -X GET "https://config.tv.pandavideo.com.br/vz-163a34ba-25b/{video_external_id}-ai.pt-BR.json"

# v1 (legacy) — no language suffix
curl -X GET "https://config.tv.pandavideo.com.br/vz-163a34ba-25b/{video_external_id}-ai.json"