Presentia
getescriba.com
- Free plan
- Web
- API
- Self-hosted
7.1 Club score Visit getescriba.com
What the maker publishes
- Brand controls
- Yes source
- Cost tracking
- The cost panel attributes model calls to a presentation, stage, and slide and uses a versioned pricing catalog. source
- Data charts
- For CSV, TSV, or JSON datasets, chart figures are checked against exact data values or column aggregates; a slide is rejected if a retry still produces incorrect figures. source
- Dataset limit
- The documented default row cap for the charts-from-data endpoint is 200 rows. source
- Deployment
- The app is packaged in a Dockerfile and can be deployed on platforms that build from a Dockerfile; persistent app data is stored under /app_data. source
- Document input
- It accepts PDFs, Word and PowerPoint files, spreadsheets, images, and plain text as source material for decks. source
- Document-to-deck generation
- Yes source
- Escriba integration
- Presentia can delegate document conversion to Escriba for OCR and falls back to local parsers if Escriba is unavailable or fails. source
- Exports
- The maker’s overview says presentations can be exported to PPTX or PDF. source
- Intended users
- The maker identifies document-to-slide users, analysts and accountants needing data-grounded charts, Markdown writers, and self-hosters or teams choosing their own models as intended users. source
- License and maintainer
- Presentia is released under Apache License 2.0 and is maintained by Diego Parras as part of the Escriba family. source
- License and maker
- Presentia is released under Apache License 2.0 and is maintained by Diego Parras as part of the Escriba family. source
- Markdown editing
- Markdown sections split by horizontal rules or headings become cards, with Preserve, Condense, and Generate text modes and a built-in editor. source
- Markdown input
- Markdown sections split by horizontal rules or headings become cards, with Preserve, Condense, and Generate text modes. source
- Model cost tracking
- The cost panel logs model calls by presentation, stage, and slide, and lets users review deck costs and compare providers. source
- Model integrations
- Supported model options named by the maker include OpenAI, Anthropic, Google, DeepSeek, OpenRouter, Ollama, LM Studio, LiteLLM, Azure OpenAI, AWS Bedrock, Google Vertex, and OpenAI-compatible endpoints. source
- Model providers
- It supports OpenAI, Anthropic, Google, DeepSeek, OpenRouter, Ollama, LM Studio, LiteLLM, Azure OpenAI, AWS Bedrock, Google Vertex, and OpenAI-compatible endpoints. source
- Notable limit
- The documentation sets DATASET_MAX_ROWS to 200 by default for the charts-from-data endpoint. source
- PDF export
- Yes source
- PII protection
- With the optional Anonimal integration enabled, content is anonymized before reaching the LLM provider, and generation stops if the anonymizer is unavailable. source
- PPTX export
- Yes source
- Privacy option
- With Anonimal enabled, prompts and extracted document text are anonymized before reaching the LLM provider, and generation stops if the anonymizer is unavailable. source
- Prompt-to-deck generation
- Yes source
- Prompt to deck
- Users can review a generated outline, choose a template, and create a presentation with layouts, text, and images in an editor with an AI assistant. source
Checked against the maker's pages on 12 Oct 2026.