💰 Token Costs
🎯 INITIAL FREE TOKENS
All users receive 1000 free tokens upon registration to try NeuralAI APIs.
💬 Chat & Conversation
10🪙
Base chat
20🪙
Chat with documents (RAG)
up to 3 attached files
🔍 Search & Analysis
20🪙
Seek Speed
40🪙
Seek Normal
70🪙
Seek Large
80🪙
Deep Research
🤖 Neuralai Agent AI
15🪙
Agent run (app & Agent API)
15 base · 30 if the knowledge context exceeds 40,000 characters · +5 per message with dynamic behavior
40🪙
🧊 3D modeling from agent (generation)
40 tokens per model: OpenSCAD + CadQuery generated from the prompt, STL files (3D printing), STEP (professional CAD), GLB and preview + editable sources · automatic refund if generation fails. Complex models generated in multiple modular components: +20 tokens per extra component (max 6 components, up to 140 tokens)
5🪙
🧊 3D modeling (model editor)
5 tokens per render: regenerate the model files after editing the source code (OpenSCAD/CadQuery) in the result editor · no AI model involved · automatic refund if the render fails
5🪙
🧊 2D CAD sections (plans, sections, elevations)
5 tokens per AI-generated section in the 2D Sections editor (plans, vertical sections, elevations) · refunded if generation fails · DXF, DWG and PDF export included · DWG conversion and saving sections charge no tokens
30🪙
📐 Engineering Calculations (deep calculation)
30 tokens per calculation (15 base + 15 deep): ULS/SLS load combinations, checks per combination, fatigue checks (S-N curves, Palmgren-Miner damage) and up to 8 plots · automatic refund if the calculation fails
10🪙
Agent prompt copilot
per generation: general prompt, welcome message and agent summary
🪙 0
Agent management via Agent API
create, update, delete and history read (agents scope) — no tokens
5🪙
Word export (.docx) from agent
5 tokens per Word document of the agent result · native headings, lists and tables · from the app or via Agent API
3🪙
Excel export (.xlsx) from agent
3 tokens per file: each markdown table becomes an Excel sheet with bold header and real numbers · from the app or via Agent API
🎨 Content Generation
50🪙
Code Generation
🪙 40-90
Generate Image
3 qualities · 512 (~0.25 MP): 40 · 1K (~1 MP): 60 (default) · 2K (~4 MP): 90 · image_quality parameter or automatic prompt detection
🪙 250-825
Generate Video
base 250 (4s) · 720p: 250 (4s), 350 (6s), 450 (8s), 550 (10s) · 1080p ×1.5 (375/525/675/825)
🪙 60/minute
Generate Music (Lyria 3)
60 tokens per requested minute, minimum 60 · Clip ≤30s, Pro up to ~3 minutes
5🪙
Professional PDF Export
5 tokens per document · A4 with cover, tables, code, sources and page numbering · the 5 tokens cover the storage credits of the generated PDF
🪙 90 / 10 pages
16:9 PDF Presentation (AI slides)
90 tokens per 10 pages · multi-theme AI deck (aurora, neon3d, lightglass, sunrise) with 3D graphics, cover and sources slide · under 10 pages by default, up to 20 on explicit request · covers generation and hosting
🎬 Video Analysis (RAG)
20🪙
Video RAG – 0-10 min
video analysis up to 10 min
35🪙
Video RAG – 10-20 min
video analysis 10-20 min
70🪙
Video RAG – 20-30 min
video analysis 20-30 min
120🪙
Video RAG – 30-60 min
video analysis 30-60 min (max)
📚 Knowledge Base
🪙 10/minute
🎵 KB Audio Transcription
MP3, WAV, OGG - multilingual
🪙 15/minute
🎬 KB Video Transcription
MP4, MOV, AVI - multilingual
🪙 5 + 1/1k char
🔗 KB YouTube (subtitles)
or 15t/min AI fallback
🪙 5 + 1/2k char
🌐 KB Web Scraping
extraction from URL
5🪙
KB Upload — TXT/DOCX/ODT (local)
no LLM, local parsing
10🪙
KB Upload OCR — 0–10k char
small PDF/images
15🪙
KB Upload OCR — 10k–50k char
~100 standard PDF pages
30🪙
KB Upload OCR — 50k–150k char
~300 pages
60🪙
KB Upload OCR — 150k–500k char
large document
100🪙
KB Upload OCR — 500k–1M char
massive document
150🪙
KB Upload OCR — oltre 1M char
huge document
🪙 15 / 30
Knowledge Base - Query
base; 30 with context over 40,000 characters
🗄️ Enterprise Database (Text-to-SQL)
2🪙
Text-to-SQL Query (generation)
per SQL query generated from natural language
💾 Storage
🪙 50 / GB per month
Extra space beyond 1 GB
1 GB free included · automatic charge from the wallet when the quota is exceeded (VAT excluded) · renewed on the 1st of each month · deleting files frees space instantly
🔌 Canvas Site Integrations
1🪙
Webhook received by the site
for each PayPal/Stripe/generic event received by the site relay
2🪙
Event forwarded to an endpoint
HMAC-SHA256 signed and sent to your backend
5🪙
Relay API call
e.g. Stripe checkout creation from the published site
🧠 AI Orchestrator
3🪙
Orchestrator (Intent Detection)
LLM intent classification + multi-turn context concatenation
🪙 +1
Conversation context (memory)
flat surcharge on each request with active memory (session_id or conversation_history)
5🪙
Dynamic behavior (dynamic_mode)
per request with dynamic_mode: true — mood/technical level/format analysis + adaptive instructions in the system prompt; charged only on successful responses (see /api-docs/dynamic-mode)
🎤 Voice & Audio
🪙 1 / 300 char
Text-to-Speech (Neural2)
Neural2 voices
Token Packs & Quantity Multiplier
• Dynamic packs and prices configurable by the administrator — visible in the API Console Tokens section
• Multiplier 1-10×: buy 1 to 10 copies of the same pack in a single transaction. Tokens and price multiplied automatically.
• Example: 3× pack = tokens and price automatically multiplied by 3.
• Available for one-time purchase and monthly auto-renewal (Stripe subscription).
• Subscription management: change quantity (1-10×) or cancel from the 'My Subscriptions' panel.
📊 How Tokens Work
• Tokens are consumed based on operations performed
• Audio/video transcriptions are billed per minute
• OCR is billed based on the number of characters extracted
• Active conversation context (session or history) adds a flat +1 token surcharge per request
• A4 PDF export (generate_pdf: true) costs 5 tokens per document; the 16:9 presentation (pdf_format: "presentation") costs 90 tokens per every 10 pages — AI deck of up to 20 pages (beyond 10 pages the cost doubles) — multi-theme deck (aurora, neon3d, lightglass, sunrise) with 3D graphics, no AI images