Pricing

Start free. Scale on your terms. From $5/month.

Name: AgenTorQ
Brand: AgenTorQ
Price: 5 USD
Availability: InStock

One governed, multi-model enterprise AI platform — every plan includes both deployment editions and every feature. Personal for individuals, graduated Team pricing that rewards scale, and custom Enterprise. Every plan starts with a 30-day free trial. Bring your own model keys and pay AI providers directly at cost.

Start 30-day free trial Talk to sales

30-day free trial on every plan Both editions & all features BYOK — no markup on AI Volume discounts as you scale Cancel anytime

Simple, all-inclusive pricing

Simple, transparent pricing. Every feature in every plan.

Two ways to buy — self-serve Individual & Team with graduated per-seat pricing, or custom Enterprise. Plans differ only by seats and scale, never by features. Every plan starts with 30 days free.

All prices in USD and exclude tax — local VAT/GST is calculated and added at checkout.

Individual

The complete platform, sized for one practitioner. Every capability the largest rollouts run on — nothing held back for a higher tier.

$5/ month

1 licence · everything included

One licence per account. Need a second person? Enterprise Teams starts at 2 seats.

30-day free trial · no card to start Start 30-day free trial

Role-aware AI Workmates with an Intent Router on every turn
BYOK across 14+ model providers — pay AI at cost, no markup
RAG over your records and files — every answer cites its sources
Voice, 3D avatars and DevSpace Studio, all included

See all 21 features

Agents

Role-aware AI Workmates with an Intent Router on every turn
218 ready templates across 16 divisions
Workmate teams — specialists collaborate on one task
Ask, Edit, Plan, Auto and Bypass modes, enforced server-side

Models

BYOK across 14+ providers — pay AI at cost, no markup
Routing, automatic fallback and latency benchmarking
Token-by-token streaming

Knowledge & research

RAG over records, files and docs — every answer cites its sources
Knowledge Bridge — include/exclude scoping per source
Enterprise search plus dated, sourced web search
PDF, Word, Excel and CSV ingestion with vision and OCR
Per-user memory — concepts, entities and workstreams recalled

Data & actions

400+ app connectors over per-user OAuth
Postgres, MySQL, Snowflake and BigQuery
REST, GraphQL, custom MCP servers and webhooks
ACTs — every write proposed on a card you approve first

Experience

Voice — every Workmate speaks and listens, hands-free
3D animated avatars with presence and gestures
Light, dark and accent themes
DevSpace Studio IDE — Apex, LWC, Flows and real git
Customer conversations free and unlimited, never metered

Your bill doesn’t move when your volume does.

Metered copilots charge for the AI seat and again for every customer conversation. AgenTorQ charges for the seat and stops there — customer-facing agents are free at any volume. Drag the sliders; holding a quote? Drop its rates into the two greyed fields. We don’t publish anyone else’s prices.

People on your team 20

Customer conversations / mo 10,000

Their AI seat add-on / user

$

How your quote bills usage

Their charge per conversation

$

Credit-based quotes bill per action, not per conversation — one conversation is usually several actions. Switch the selector to match the quote you are holding, then set both figures from it.

A metered platform would bill ~~$22,500~~ a month

You pay just

$500

98% less than a metered platform

per month · 20 seats, everything included

$0 per conversation
Customers free & unlimited
AI on your own key, at cost

You save

Metered platform: ~~$270,000~~/year → AgenTorQ: $6,000/year

$264,000/ year

stays in your business — 45× lower

That is 44 years of AgenTorQ, paid for by one year of not being metered.

What happens as you grow

Monthly cost against customer conversations, using the rates above. One line is a slope. The other is a straight, flat line — because conversations are free.

Metered platformEvery conversation adds cost. The line never flattens.

AgenTorQSeats only. Volume changes nothing.

Model tokens are billed to you directly by your AI provider and are not shown here — they apply to any platform you choose.

Same team of 20. Only the conversation volume changes — your column never moves.
Conversations / mo	A metered platform	AgenTorQ	You save / year

How the models differ

Metered by conversation, or flat by seat.

A native, single-vendor CRM agent charges for the AI seat and again for every customer conversation. Both lines grow when your business does. AgenTorQ charges for the seat and stops there.

	Native single-vendor CRM agent	AgenTorQ
External / customer users	Licensed and metered	Free. Unlimited. No seat, no charge.
Customer conversations	Billed per conversation or per action	Unlimited. Never metered.
AI model	One vendor’s model family	14+ providers, your own keys, switch any time
Who bills you for AI	The platform, with its own margin	Your provider, at their cost — zero markup
Cost as volume grows	Rises with every extra conversation	Flat. Only your seat count changes it.
Token visibility	Aggregate credits, after the fact	Per run, per agent, per user — plus a forecast
Where it runs	Inside one CRM	Native Salesforce package and standalone web — both included
Staff and customer agents	Separately licensed and separately metered	Both surfaces, one plan

Two ways to deploy · one platform

Both editions are included in every plan.

AgenTorQ has two deployment editions of one engine. Run the standalone web workspace, install the Salesforce-native package, or use both — the price and the feature set are identical.

Edition · Standalone web

AgenTorQ Cloud

A standalone AI workspace that connects Salesforce, Microsoft 365, Google Workspace, Slack, GitHub, databases and APIs — no CRM required, and nothing to install.

Included in every plan · start instantly, no install

Instant browser workspace — sign in and go
Universal connectivity across your business apps
Per-user OAuth; personal chats stay private
Centralized governance & feature-access RBAC

Launch Workspace

Edition · Salesforce-native

AgenTorQ for Salesforce

A real AppExchange managed package that installs inside Salesforce — LWC, Apex, Flows, record pages and the utility bar — enforced in USER_MODE, admin-managed, with data staying in Salesforce.

Included in every plan · native, admin-managed install

Native managed package on the AppExchange
Record-aware AI on every page & the utility bar
Salesforce sharing, CRUD/FLS & USER_MODE enforced
Live CRM grounding with SOQL/DML guards

View AppExchange Edition

Same AI engine · Same Workmates · Same governance · Same connectors · Same model routing. One platform, every plan.

Everything included

One price unlocks the whole platform.

No add-ons and no feature gates. Every capability of the enterprise AI platform ships in every plan, to every user — explore each in depth on its own page.

AI Workmates

Role-aware agents for Sales, Service, Ops, Admin and Developer that read data, reason over knowledge and take approved actions — plus 218 templates.

Intent Router

Every turn is graded and routed to data lookup, knowledge, reasoning or a real action — nothing hardcoded.

Multi-model routing

Bring your own keys across 14+ providers — OpenAI, Claude, Gemini, Grok, DeepSeek, Mistral, Groq and more — with cost control and fallback.

RAG with citations

Secure retrieval over folders, files and chunks with embeddings — grounded answers that cite live records and documents every time.

400+ connectors

Salesforce, Microsoft 365, Google, Slack, GitHub and hundreds more via per-user OAuth, plus REST, GraphQL, databases and custom MCP.

Workflow actions

Create and update records, run Flows and send email behind approval cards — workflow execution, not just chat.

Governance & audit

RBAC, named PII masking, SOQL/DML guards, USER_MODE in Salesforce, action approvals and an immutable audit trail.

DevSpace Studio

An AI-native VS Code-style IDE for Apex, LWC and Flow with GitHub branch, PR and merge — ask, plan, diff, approve, deploy, test, fix.

Document & OCR

Read and reason over PDF, Word, Excel, CSV and scanned images with native vision or OCR.

Per-user memory

Each Workmate keeps private Concepts, Entities, Workstreams and Notes with cross-session recall — isolated per user.

Web search

Model-agnostic web search with dated, sourced results, alongside a reusable prompt library.

Workspaces & RBAC

Invite users, build SFDC-style profiles from a feature-toggle grid, and enforce deny-by-default access server-side.

Explore Platform Explore Connectors View Security

AI at cost · BYOK

Pay for the platform. Pay AI providers at cost.

Vendor-locked copilots meter you per conversation and mark up every token. AgenTorQ separates the two: a predictable flat platform fee, and your own model keys billed directly by the provider.

Flat, predictable platform fee

One simple platform fee covers everything — both editions and every feature. Volume discounts as you scale, and no surprise usage bill.

Your keys, your cost

Connect your own provider keys and pay OpenAI, Anthropic, Google, Mistral, Groq and others directly at their published rates — AgenTorQ adds no markup.

Control the model spend

Route by task, cost and latency, set fallbacks, and restrict providers per role — so teams use the right model without runaway spend.

No AI seat tax · No per-conversation credits · No per-message metering. Just one simple platform fee.

Transparency, built in

You can watch every step, and every token.

Most AI products give you a spinner and a surprise invoice. AgenTorQ shows the agent thinking step by step while it works, meters every run as it happens, and forecasts where the month’s spend is heading before the bill lands.

Think time, in the open

Each step renders as the agent plans, retrieves and calls tools — a live row for the step running now, a tick and a duration for each one done. When a turn is slow, you can see which hop was slow.

Token meter, per run

Every run’s token usage is metered and attributed per user, model and task type, with an estimated cost from the price you set. Spend is visible as it accrues — not discovered on the invoice.

Spend forecast

A statistical forecast projects where the month’s spend is heading from your live daily usage and model mix — you see the trend early, and it reports how much of the volume was priced so an estimate is never mistaken for the whole bill.

Estimate your model spend

A quick projection of what your provider will bill — the same inputs the in-app dashboard tracks. We don’t resell model usage, so we have no reason to hide any of it; every assumption below is yours to change.

Model tier

Input, per million tokens$ Output, per million tokens$

Your assumptions

Tokens in, per interaction Tokens out, per interaction Team interactions per person / mo

Prompt caching on — repeated context bills at about a tenth

Estimated AI spend

$0

per month, paid to your provider

Interactions / month0
Tokens / month0
Cost per interaction$0.00
Per year$0

Paid directly to your AI provider at their published rates. AgenTorQ adds no markup, takes no cut, and never sits between you and your key.

Straight answers

The things people ask before they buy.

No usage meters, no per-conversation credits, no markup on AI — and your customers cost nothing. The questions that come up most, answered plainly.

Do you charge per message or per conversation?

No. There is no usage meter of any kind. You pay the flat plan price, and your AI provider bills you directly for model tokens at their published cost. We add no markup.

Are customer-facing agents really free?

Yes. External and customer users are unlimited and carry no seat cost and no conversation cost. Paid seats cover your own team only.

How does graduated Team pricing work?

It is marginal, so you only pay the higher rate on seats in that band. Seats 1–10 are $30, 11–30 are $20, 31+ are $10. A 15-seat team is 10×$30 + 5×$20 = $400 a month.

Do I pay twice for Salesforce and the browser edition?

No. They are the same platform and both are included in every plan. Run either, or both, with an identical feature set.

What happens when my volume doubles?

Nothing changes on your invoice. Your model provider will bill more tokens, and that is the only line that moves. Your plan price is driven by seat count alone.

Is there a free trial?

Every plan starts with 30 days free. No card up front, and you can cancel any time before it ends at no cost.

Which models can I use?

Any of 14+ supported providers, including open-source and self-hosted endpoints. Your keys stay yours, and callouts go from your org straight to the provider.

Can I get an MSA, DPA or SLA?

Yes, on Enterprise, alongside SSO/SAML, a security and architecture review, multi-org rollout and dedicated support.

Enterprise & volume

Rolling out across the organization?

Team pricing is graduated, so it already rewards scale as you add seats. For custom volume licensing, SSO/SAML, an SLA, procurement and legal (MSA/DPA), a security and architecture review, or a multi-org rollout, talk to us for Enterprise terms.

Book Demo View Security

Volume licensing Procurement, MSA & DPA Security & architecture review Compliance-ready architecture Multi-org / multi-team rollout Onboarding & priority support

Pricing FAQ

Pricing questions, answered.

Everything about our plans, graduated Team pricing, the 30-day free trial, and bring-your-own-key AI.

How much does AgenTorQ cost?

Three plans, all with the same features. Personal is $5/month for 1 user. Team is graduated volume pricing — the first 10 seats are $30 each, seats 11–30 are $20 each, and seats 31+ are $10 each per month (a 15-seat team is 10×$30 + 5×$20 = $400/mo). Enterprise is custom. Every plan starts with a 30-day free trial.

How does the graduated Team pricing work?

It is graduated (marginal), so you only pay the higher rate on the seats in each band: seats 1–10 at $30, seats 11–30 at $20, seats 31+ at $10. Examples: 5 seats = $150/mo, 15 = $400/mo, 30 = $700/mo, 50 = $900/mo, 100 = $1,400/mo. The total always rises with more seats, while the effective per-seat cost falls as you scale.

Is there a free trial?

Yes — every plan starts with 30 days free. You are not charged during the first 30 days and can cancel anytime before it ends at no cost.

Does AgenTorQ charge for AI usage or per message?

No. AgenTorQ is bring-your-own-key (BYOK) across 14+ model providers. You connect your own keys and pay OpenAI, Anthropic, Google, Mistral, Groq and others directly at their cost. AgenTorQ adds no markup on AI and never meters per message — you pay only the flat plan price.

What is included in each plan?

Every plan includes both editions, AI Workmates, the Intent Router, multi-model BYOK routing across 14+ providers, secure RAG with citations, enterprise search, 400+ connectors plus REST/GraphQL/database/MCP, workflow actions with approvals, RBAC and PII masking, an immutable audit trail, per-user memory, the prompt library, document and OCR intelligence, and DevSpace Studio for Apex, LWC, Flow and GitHub. There is no feature gating — only seat count and scale differ.

Do I pay separately for the Salesforce edition and the web edition?

No. Both editions are the same platform and both are included in every plan. Deploy AgenTorQ Cloud as a standalone web workspace, install AgenTorQ for Salesforce from the AppExchange, or run both — the feature set is identical.

Is there enterprise or custom pricing?

Yes. Enterprise is custom-priced and adds SSO/SAML, an SLA, MSA/DPA, a security and architecture review, multi-org rollout and dedicated support. For large volumes, procurement and legal, email reachout@agentorq.com and we will scope terms with you.

How do I get started?

Launch the standalone web workspace and start your 30-day free trial instantly with no install — sign in, open Subscription and pick Personal or Team. For custom or Enterprise terms, email reachout@agentorq.com.

Governed enterprise AI. Start your month free.

Start in the web workspace in seconds, or book a guided walkthrough of either edition. One platform, every plan — you choose where it runs.

Launch Workspace Book Demo

reachout@agentorq.com