Athena
CoreFoundational Intelligence
The essential brain add-on for every agent. Delivers real-time time/date awareness and safe basic calculations to ground reasoning and prevent hallucinations.
$0.00075 per call
INFERENCE · INCLUDES ALEXANDRIA
PRODUCTION AI INFERENCE.ELITE MANAGED TOOLING.
OpenAI-compatible inference — elastic models or dedicated endpoints — on one prepaid key. Included: Alexandria, the library your agents actually run. Enable once; available on Chat, the API, MCP, and Chronos. Atlas is coming.
Elastic or dedicated inference on one key. Alexandria ships with it — named tools, same wallet.
AthenaOlympusVesperHeraldAtlasSOON
AthenaOlympusDeskAtlasSOONChronosVesperHeraldPandoraIrisLedger
import OpenAI from 'openai';
const client = new OpenAI({
baseURL: 'https://hypervize.tech/api',
apiKey: process.env.HYPERVIZE_API_KEY,
});
const response = await client.chat.completions.create({
model: 'google.gemini-3.7-flash' // or 'auto' — we pick the best model by default,
messages: [{ role: 'user', content: 'How deep is the Marianas Trench?'}],
max_tokens: 1024
});
// Enabled Alexandria tools inject on this same call
// Athena · Vesper · Herald · Ledger — same key, no second API
COMPUTE → REGISTRY → ENDPOINT
FROM TRAINING TO INFERENCE
- 01Fine-tune on Hypervize Compute — provisioned GPUs or HGX.Provisioned compute →
- 02Weights land in the Hypervize Registry. No USB-stick ritual.
- 03One-click dedicated endpoint. Same OpenAI-shaped API as elastic.Dedicated docs →
ZERO-CONFIG NETWORKING
BRING YOUR OWN DOMAIN.
Add a CNAME. We provision TLS and route to your dedicated endpoint. Auto *.hypervize.tech if you do not bring a name.
Dedicated docs →CHOOSE YOUR DEPLOYMENT ARCHITECTURE
BEST FOR STANDARD MODELS
Elastic Inference (Burstable)
Priced per 1 Million Tokens
Curated frontier models on our shared cluster. No cold starts. You pay for the tokens you generate — not a junk catalog of a thousand IDs.
- ▸ Auto-scaling concurrency
- ▸ Fully managed
- ▸ Global rate limits apply
ENTERPRISE GRADE
Dedicated Endpoints
Priced per Hour (Compute Based)
Private GPU — fractional or full — with your fine-tune or any Hugging Face model. BYOD. Scale to zero. Throughput is yours, not a noisy neighbor's.
- ▸ Custom weights
- ▸Guaranteed SLA & throughput
- ▸ Unlimited tokens (hardware-bound)
Dedicated resources billed weekly based on usage (min 1h if used) + full-week addons via saved PM. Fine print on docs.
THE ALEXANDRIA PROJECT
KNOWLEDGE * MEMORY * ACTION
THE LIBRARY INSIDE THE PIPE
Enable a tool once. It is available on Chat, the API, MCP, and Chronos — same key, same prepaid ledger. Atlas is coming.
In Chat, the same library is the tools panel — not a second product.
Olympus
CoreAutomatic Model Choice
Hypervize picks the right model for you — every time. On by default. No charts, no second-guessing Claude vs GPT vs Grok. Lock a model anytime.
Free
Desk
CoreYour Computer
Let Chat work with files on a computer you connect — Documents, Desktop, Downloads, Pictures. Pair in Chat. Saving, deleting, or running a command always asks first.
$0.015–$0.03 per call
Atlas
COMING SOONEternal Memory
The foundation of persistent agents. Store facts, histories, embeddings, and structured knowledge that survives every conversation, schedule, and session.
Storage-based • $0.002/MB + queries
Chronos
AutomationAutonomous Time
Enable scheduled tool and model recipes. When on: Schedule appears on chat replies that used tools, and API jobs work with your key. $0.01 per run plus tools and any model tokens.
$0.01 per run + tools + tokens
Vesper
ResearchWeb Intelligence
Full internet browsing for agents. Search or fetch any page and extract exactly what you need with natural language instructions. Supports structured output (JSON tables, metadata, headings, etc.).
$0.015 per call
Herald
DeliveryDirect Agent Voice
Send plain text, Markdown, or HTML emails to any address — only when you explicitly ask. Optional Library attachment via attachment_file_id (private docs loaded server-side). 20 emails/hour, max 10k char body.
$0.05 per call
Pandora
ActionsOpen the Box
Full read and write access across your email, calendar, and file services. Open Pandora’s box and unleash your agents into your real digital world.
$0.015 per call
Iris
CreationImage Creation
Turn language into images. Describe what you want; Hypervize selects Imagine Image, Imagine Image Quality, or Imagine Image 2.0 (precise). Pay per image — not a monthly creative suite.
$0.015 per call
Ledger
FilesFiles & Data
Create, read, and convert files (CSV, JSON, PDF, text). Stored privately.
$0.03 per call
- 01 ENABLEIn Alexandria (dashboard), or POST /api/tools with your key.
- 02 CALL/api/chat/completions or Chat. MCP execute is gated the same way; Chronos needs its own toggle.
- 03 RUNTools run server-side. Billed as tool calls on the prepaid balance.
ELASTIC MODEL PRICING
Tokens and Alexandria tool calls draw from the same prepaid balance. Dedicated weekly invoices are GPU time only. Billing docs.
| Model Name | Context Window | Input (1M) | Output (1M) |
|---|---|---|---|
Gemini 3.1 Pro google.gemini-3.1-pro | 1M | -- | -- |
Kimi K3 moonshot.kimi-k3 | 1M | -- | -- |
Claude Fable 5.1 anthropic.claude-fable-5-1 | 1M | -- | -- |
Grok 4.6 xai.grok-4.6 | 500k | -- | -- |
GPT-5.6 Sol openai.gpt-5.6-sol | 1M | -- | -- |
GPT-5.5 openai.gpt-5.5 | 1M | -- | -- |
GPT-6 Astra openai.gpt-6-astra | 1.05M | -- | -- |