Question 1

How is Orion different from GPT-5, Claude, or Gemini?

Accepted Answer

Frontier models can describe what they see, but not act on it. Orion goes beyond perception by planning, executing, and validating visual tasks. Instead of just describing an image, Orion can detect, crop, segment, generate, and analyze in sequence, reliably and deterministically.

Question 2

What can I build with Orion?

Accepted Answer

Developers use Orion for visual ETL pipelines, automated product and marketing asset generation, document parsing and redaction, video summarization and clipping, medical and geospatial visual analysis, UI testing, and more.

Question 3

How accurate is Orion compared to traditional CV models?

Accepted Answer

Orion combines modern VLM architectures with traditional computer-vision models. In benchmarks across multi-modal tasks, Orion consistently outperforms leading VLMs on multi-step visual reasoning, structured OCR, and traditional CV tasks.

Question 4

Can VLM Run handle handwritten and scanned documents?

Accepted Answer

Yes. Orion reasons visually about layout and structure, not just flat text. It handles messy scans, faxes, handwritten forms, and complex table layouts that traditional OCR struggles with.

Question 5

How is Orion priced?

Accepted Answer

Credits start at $0.01 each (100 credits = $1) with 1,000 free sign-up credits. See https://vlm.run/pricing for tier breakdowns and per-domain rates.

Question 6

Is there a free trial?

Accepted Answer

Yes. Every account starts with 1,000 free credits ($10) and no card required. Upgrade to Pro or Enterprise whenever you need more throughput or compliance.

Question 7

Do unused credits roll over?

Accepted Answer

Pro plan credits are replenished monthly and don't roll over. Overage credits are billed on demand at the standard rate.

Question 8

When should I use visual grounding?

Accepted Answer

Use grounding when you need an audit trail. Grounding returns bounding boxes and confidence scores for every extracted field, required for healthcare, finance, and legal QA workflows.

Question 9

Can I mix service tiers within one project?

Accepted Answer

Yes. Set the tier per request so real-time user calls can use priority while nightly batches use flex, all from the same API key.

Question 10

How does billing work for multi-page PDFs?

Accepted Answer

Each page is billed independently at the configured detail level. A 20-page PDF at auto = 2 × 20 = 40 credits ($0.40). Grounding adds +2 per page.

Question 11

What's included in Zero-Data Retention?

Accepted Answer

With ZDR enabled (Pro and Enterprise), your files and API responses are never persisted on our infrastructure. Processing happens in memory and results stream directly back to you.

Question 12

What payment methods do you accept?

Accepted Answer

We bill through Stripe (Visa, Mastercard, American Express). Enterprise customers can arrange invoice billing with custom net terms.

Question 13

What types of documents do you support today?

Accepted Answer

Common healthcare documents including referrals, intake forms, insurance cards, faxes, claims packets, and EOBs, structured for EMR/EHR and FHIR workflows.

Question 14

Can you handle faxed or low-quality scanned documents?

Accepted Answer

We handle scanned PDFs, faxed referrals, handwritten or typed forms, insurance card photos, and multi-page claims. If we don't support something yet, we can often add it in under 24 hours.

Question 15

Is this production-ready for healthcare?

Accepted Answer

Yes. HIPAA-ready deployments, private VPC hosting, audit logging, and schema-level validation for regulated workflows like referrals and claims.

Question 16

How is this different from OCR or generic LLM APIs?

Accepted Answer

Text-based OCR and generic LLM APIs primarily read text. VLM Run uses vision-language models to reason about layout, structure, and relationships, then enforces deterministic healthcare schemas on top.

Question 17

Can VLM Run handle low-quality scans and blueprints?

Accepted Answer

Yes. Our models handle faded scans, hand-marked drawings, low-res PDFs, and on-site photos. Orion reasons visually about layout and structure, not just characters.

Question 18

What document types does VLM Run support?

Accepted Answer

Blueprints, floorplans, schedule tables, BOQs, material lists, submittals, RFIs, and spec sheets including merged cells, nested headers, and multi-page tables.

Question 19

How does VLM Run compare to traditional OCR for construction documents?

Accepted Answer

Traditional OCR fails on complex table layouts, drawing symbols, and engineering annotations. VLM Run uses vision-native models trained on technical documents to reason about structure.

Question 20

Is my project data secure?

Accepted Answer

VLM Run is SOC 2 Type II certified. Enterprise customers can use VPC deployment so data never leaves your cloud. All data is encrypted in transit and at rest.

Question 21

How do you keep data private?

Accepted Answer

Pro-tier runs in our private cloud with API logging in the observability dashboard. Enterprise-tier supports SOC 2, GDPR, and HIPAA with configurable retention including ZDR.

Question 22

Can I run Orion on-prem or in-VPC?

Accepted Answer

Yes. Enterprise deployments support VPC Peering and In-VPC hosting so data never leaves your environment. SOC 2 Type II and HIPAA-ready.

Question 23

Can I upgrade or downgrade my plan anytime?

Accepted Answer

Yes. Changes take effect immediately with prorated billing. Enterprise plans include invoice billing and custom SLAs.

FAQ

Product & Orion

Pricing & Billing

Healthcare

Construction

Security & Enterprise

Still have questions?