Find answers to the topics that matter to you here
Veontra is an AI workspace for documents, audio, images, video, music and 3D. It extracts structured data from invoices, receipts and other files, chats with PDFs with page citations, transcribes audio, turns text into speech, and generates images, video clips, songs and 3D models — plus an optional desktop Copilot overlay. All in one dashboard, with human review before export where it applies.
Veontra works for personal use as well as for teams and organizations. It is especially useful in accounting and finance, logistics, retail, legal services, recruiting (HR), and content teams — anywhere teams process documents or create media regularly.
Sign up as an individual or as an organization — you get a 14-day trial and 60 credits, once per IP address. No credit card required.
Veontra Copilot is a Windows desktop overlay (macOS coming soon) that stays on top of your other apps. Capture screenshots, dictate with voice, and chat using the same Veontra account and credit balance as the website. Download it from the Desktop Copilot page after you sign in.
You upload a PDF or image, AI extracts the fields, you review and approve the values in the app — then the data is ready to export. The whole cycle takes minutes.
Veontra never finalizes results automatically. Every document goes through human-in-the-loop review: you edit fields and confirm them before export, so you stay in control of accuracy.
Yes. In addition to ready-made templates, you can create a custom field schema for your document type and edit fields on presets.
Yes. Beyond field extraction, Veontra includes document chat, audio transcription, text to speech, AI image generation, restyle and edit, AI video generation, AI music generation and AI 3D model generation — plus a desktop Copilot overlay. They share the same dashboard and credit balance.
Ready-made templates are available for invoices, receipts, resumes, bank statements, and ID documents, plus a custom schema mode and a text mode for any other documents.
You can upload PDFs and images (such as photos or scans). Multi-page PDFs are supported too.
Use custom schema mode: define the fields you need — and Veontra will extract exactly those. For free-form text, use text mode.
Upload a PDF (up to 100 pages) and ask questions in plain language. Answers come from that file only — not the open web — and can include page citations so you can jump to the source in the preview.
Credits are charged for the pages in the PDF and for message packs in the conversation. See the Credit services page for current rates.
Upload an audio file and Veontra converts speech into a timestamped transcript. Turn on speaker diarization to label who said what, or translation to get each segment in another language. Review segments side by side with a built-in audio player before approving.
MP3, WAV, M4A, AAC, OGG, WebM, and FLAC are supported. The free demo is limited to a 1-minute clip and 15 MB; full transcription in your dashboard supports much longer files.
Approved transcripts export to TXT, SRT, VTT, or DOCX, or to a new Google Docs file in your connected account. Bulk export from the list works the same way as single-file export.
Audio is billed per minute of audio (minimum one minute), with optional add-on costs for speaker diarization and translation. See the Credit services page for exact per-minute rates.
Write or paste a script, pick an OpenAI gpt-4o-mini-tts voice, and generate an MP3 voiceover. Optional style instructions control tone and pacing.
4 credits per minute of generated audio, rounded up. Affordability is checked before generation starts. See Credit services for the current rate.
Veontra supports eleven models from four providers: OpenAI GPT Image 2, GPT Image 2.5 Flare, and GPT Image 2.5 Sunburst; Google Nano Banana, Nano Banana 2, Nano Banana 2 Lite, and Nano Banana Pro; xAI Grok Imagine and Grok Imagine 2.0; and Black Forest Labs FLUX.1 Pro and FLUX.2 Pro. Pick the one that fits your prompt and budget.
Yes. Choose standard or high quality and generate up to three images per request. You can also attach a reference image so the model edits or matches its style, depending on the model you pick.
Image generation is billed per image, and the price depends on the model and quality — for example GPT Image 2 costs 13 credits at standard and 45 at high quality, while GPT Image 2.5 Flare starts at 4 credits. Generating several images in one request multiplies the cost by the count you choose. See Credit services for the full table.
Restyle keeps the subject from your source image and applies the look of a style reference — lighting, palette, clothing, atmosphere. You need both images plus a short prompt. It works with FLUX.2 Pro, Grok Imagine 2.0, Nano Banana 2, Nano Banana Pro, GPT Image 2, and GPT Image 2.5 Flare/Sunburst. Open the Image restyle page to see how it works, then use Restyle in Image → Generate.
Edit changes one photo in place: retouch a region, outpaint a tighter crop, replace the background, relight, or move the camera. You need a source image and a tool — not a second style reference. Open the Image edit page, then use Edit in Image → Generate.
Veontra offers seven video models: xAI Grok Imagine Video, Grok Imagine Video 1.5 and Grok Imagine Video 1.5 Lite; Google Veo 3.1 Lite, Fast and Standard; and Black Forest Labs FLUX 3 Video. Grok clips run up to 15 seconds; Veo clips run 4, 6 or 8 seconds; FLUX 3 clips run 5–20 seconds with native audio.
Yes. Attach a start frame so the clip begins from your image (image-to-video) — Grok 1.5 requires one. On Veo models you can also add an end frame so the clip interpolates between two stills, which is useful for controlled camera moves.
Video is billed per second of generated footage. Cost depends on the model and resolution — Grok and Veo Lite cost less per second than Veo Standard at 4K. See Credit services for the current per-second rates.
Four models: Google Lyria 3 Pro turns a text prompt — or a reference image — into a full song from 60 seconds to 3 minutes, with genre, mood, vocals, and tempo. MiniMax Music 2.0 follows an audio reference; MiniMax Music 2.6 is lyrics-first with no reference file; Sonilo v1.1 generates instrumental tracks up to 10 minutes.
On Lyria and MiniMax: choose instrumental, auto lyrics from your prompt, or custom lyrics you paste. In custom mode you can generate a lyrics draft (billed separately), then edit before creating the track. Sonilo is instrumental only and has no lyrics modes. An optional AI album cover can be generated alongside the song.
Lyria and MiniMax charge a flat credit cost per song. Sonilo is billed per second of audio. An album cover adds the image generation cost. A lyrics draft in custom mode is a separate flat charge. See Credit services for current rates.
Yes. Describe an object in text or pick an image, and Veontra generates a textured 3D model with PBR materials. You can rotate it in the browser, view it in AR on your phone and download it as a GLB file for Blender, Unity, Unreal or your website.
You can change its material: open a finished model, choose Change material and describe the new color or material — the shape stays the same and the result is saved as a new model. To change the shape, edit the source image or prompt and generate again.
Per generation: from 44 credits for a model from text and 66 credits from an image. HD texture, Ultra geometry and smart low poly add to the base price, and changing the material starts at 22 credits. See Credit services for current rates.
Yes. Images, videos, and music tracks can each be shared to the public Gallery and liked by other users, or kept private — sharing is optional and per file.
Yes. Download individual images, videos, or MP3 tracks, or grab a ZIP archive of a full generation, from your dashboard or from the Gallery.
Files are stored in a private Amazon Web Services (AWS S3) object storage bucket with time-limited signed URL access. There is no direct public access to files.
Data is isolated by organization: only you and members you invite can see it. Other users cannot access your documents.
When you delete a document, the file is removed from storage along with related data.
Veontra is built on Amazon Web Services (AWS) infrastructure and follows GDPR principles: data is encrypted in transit, isolated by organization, and deleted from storage on deletion request. Payments are processed by Paddle — we do not store your card details.
Yes — generation and extraction requests are sent to the AI provider behind the tool and model you choose (OpenAI, Google, xAI, Black Forest Labs, MiniMax and Sonilo via fal.ai, or Tripo for 3D) to process your file or prompt. We do not sell your data or use it to train AI models, and your files stay in Veontra's private AWS S3 storage.
Every action has its own credit cost. Document extraction starts at 2 credits per page at standard quality; high quality and optional field highlighting cost more per page. Images, video, music, speech, 3D and document chat each have their own rates. See the credit services page for the current price of each action.
Subscription — a chosen monthly credit volume at a lower price, from $3.49/month at 500 credits (billed every 3 months). Pay-as-you-go — a one-time credit purchase from $4.99 for 500 credits, with no commitment. Features are the same, and pay-as-you-go credits can be added on top of a subscription.
New accounts get 14 days and 60 credits to test extraction, transcription, and generation on their own files, once per IP address. No credit card required.
Images charge per image after a successful generation; video per second of footage; text to speech per minute; 3D models per generation, with options like HD texture adding to the base price. Music is a flat cost per song on Lyria and MiniMax, or per second on Sonilo. Failed generations are refunded automatically, so you are not charged for jobs that error out.
Approved documents with extracted fields can be exported to CSV, Excel, or JSON, and — when the feature is enabled — to a new Google Sheets spreadsheet in your account.
Yes. Export API is available on both subscription and pay-as-you-go with the same capabilities — you can connect it to your systems and automate exports.
At your explicit request, Veontra creates and fills a new spreadsheet in your Google account. We do not view or edit your existing files.
Text is a separate Veontra tool for extracting full text from images (JPEG, PNG, WebP). AI extracts the text, you review and approve the result, then export it to TXT, DOCX, or Google Docs — one document or several at once.
Yes. Download generated images and video clips directly as image or MP4 files, and download music as MP3 or as a ZIP archive that includes the cover art. Where Google Drive export is enabled, files can also be exported there.
Yes. The REST API v1 covers document extraction with webhooks, request logs, and API key management. See the API docs page for endpoints and code samples in cURL, Node.js, and Python.
The Export/Developer API is available on both subscription and pay-as-you-go plans with the same capabilities.
Create and revoke API keys from your dashboard's Developer settings — up to 10 active keys per organization, plus webhook configuration for delivery events.
Extract documents, transcribe and voice audio, generate music, images, video and 3D models — all from the same dashboard.
