Nous Research

What it does.

Nous Research is an AI research and product organization whose official site links Hermes Agent and the Nous Portal. Its work includes models, agents and supporting infrastructure. Because the source chart is a vendor-spending list, the entry remains Nous Research: available evidence does not allocate that spending to a particular chat product, subscription or model.

Technical users investigating open AI tools, agent workflows and the operational choices behind hosted model access.

Official product and documentation links let readers distinguish research outputs, local software and hosted services instead of conflating them.

This ranked name identifies a vendor. Its products and billing plans are described separately; a vendor-level spend rank is not a product ranking.

• Observed U.S. consumer-card spend: #50 · edition 7 · 2026-08 (article context; chart does not print a month) · United States · source label: Nous Research · vendor scope · No volume is supplied. Rank does not measure quality or suitability. U.S. consumer-card panel spending, not global revenue. August comes from article context; chart month is absent. Vendor scope does not isolate an individual app or its AI features. · Method and source

Make a small first attempt.

1. Read the linked official product information and check regional access.
2. Choose one fictional or public sample with a clear expected output.
3. Review permissions, usage limits and the applicable terms before using a feature.
4. Inspect the result manually; keep an unchanged copy of the input.

• web — verified · 2026-10-05

Understand the bill.

Pricing status: unclear · checked 2026-10-05

Status: not_found

Description: Checked October 5, 2026: Nous links a subscription-management page for Portal. Current public plan amounts and the exact composition of ranked spending were not established.

Official pricing url: https://portal.nousresearch.com/manage-subscription

Free access: Current hosted free allowances were not established; open-source software and hosted inference are separate.

Caution: Confirm current region, billing interval, renewal price, usage allowances and taxes. Annual-effective monthly figures are not cancellable monthly subscriptions.

Portal subscription route; amounts not established ↗

Know the boundaries.

• Agent tools can access files or execute actions when configured. Open-source availability does not establish model licensing, privacy, safety or the total cost of operation.

Privacy and data handling

Privacy evidence dated 2026-10-05

Privacy url: https://portal.nousresearch.com/privacy

Terms url: https://portal.nousresearch.com/terms

Status: links_only

Data to consider: Keep secrets, personal files and connected accounts out of initial experiments. Evaluate local permissions separately from the policies of any hosted model provider.

Safe use note: Keep secrets, personal files and connected accounts out of initial experiments. Evaluate local permissions separately from the policies of any hosted model provider.

Commercial use: Not established beyond the linked service and plan terms; input and third-party rights still require review.

Age or region restrictions: Not comprehensively established; check the applicable app-store region and current terms.

Vendor identity and official product links ↗ Portal subscription route; amounts not established ↗ Portal privacy, not every research product ↗ Portal terms ↗ OpenRouter official product ↗ Cursor official product ↗

IndustryNext / Practical work

IndustryNext: a bounded work scenario

Work scenario: IndustryNext: prepare an evaluation plan for an agent working only on a fictional documentation folder. A technical owner specifies read permissions, expected output and stopping conditions before execution.

Input: Fictional or already-public material, as specified in the scenario.

Workflow: Set a narrowly defined task and an expected output.

Workflow: Work only with the safe sample described above.

Workflow: Compare the result against the source and obtain human review.

Output: A reviewable internal draft or evaluation record, not an automatically published deliverable.

Human review: The role identified in the scenario verifies accuracy, rights and suitability before use.

Limits: Suggested application only; no hands-on testing or productivity claim.

Claim type: editorial_application

UES / Independent learning

UES: try an independent learning exercise

Learning goal: Practice critical evaluation using the task-specific comparison described above.

Safe sample data: Original fictional material, public documentation or an owned object photo; no private, confidential, child or regulated data.

Steps: Prepare the safe example or public documentation.

Steps: Compare the proposed output or workflow with a manual baseline.

Steps: Record one error, uncertainty or permission boundary.

Deliverable: A short annotated comparison or learning log.

Reflection: What evidence would you need before relying on this result in a real situation?

Access alternative: Use a paper storyboard, manual draft or public product documentation if access is unavailable or paid.

Institution note: Independent community learning project; no accreditation or credential is implied.

Claim type: editorial_learning_exercise

1. University of El Segundo: diagram the boundary between local agent, model provider and user files using public documentation. Identify where data leaves the device. This is independent, non-accredited community learning.

Independent learning activities; no accreditation or academic credit is implied.