A complete ecosystem, at a fraction of the cost of the alternatives
Hub Forge AI is not a self-service subscription. Every deployment is handled by our team: audit, installation, configuration and training included.
A turnkey deployment, not a self-service software
Hub Forge AI installs on your infrastructure, connects to your tools, and adapts to your industry. Every deployment includes an audit, a custom installation and training for your team.
Audit & Quote
We analyze your IT system, your tools and your needs. Quote within 48h.
Installation & Configuration
Hub Forge AI is installed on your machine or VM. Agents and connectors configured for your context.
Training & Production
Your team is trained. The Hub is operational. Ongoing support included.
A one-time investment at launch
Audit, installation, configuration and training. The Pilot tier is deductible from your first license year.
Pilot
Deductible from your first license year
- 1 expertise center, 3 agents
- 1 MCP connector
- Half a day of training
Standard
Complete deployment for your team
- 2 expertise centers
- Connectors on your key tools
- Business rules (FORGE.md)
- 1 day of training
Custom
Scope defined with you
- 3+ expertise centers
- Advanced integrations
- Reinforced Pre-Hook governance
- Dedicated support
À la carte modules
A base at 99 EUR/month, then each module has a fixed monthly price. Add or remove modules at any time. Annual commitment billed monthly, or −15% if paid upfront.
Platform
The base of your solution
- Central Hub + authentication
- Agent Monitor (real-time supervision)
- 1 expertise center of your choice
- 3 agents included
- Cloud deployment
- Standard support
Expertise center
Cyber, Ops, Compliance, Data, B2B Sales
- Curated open source GitHub skills
- Domain-specialized agents
- Automatic continuous monitoring
- ForgeGuard (CVE analysis)
- AI required
Audit framework
MSP, IT services firm (ESN), CIO, MSSP, Accounting, Recruitment
- Smart questionnaire
- Real-time scoring
- Critical gap detection
- Correlations with the Forges
- Auto AI quote and report (optional)
- Works even without AI
LeadForge
B2B prospecting — Golden Records
- Golden Records: decision-maker + verified email
- Real-time business registry — natively integrated
- Automatic resolution of the real website
- SMB/mid-market filter — zero noise, zero sole trader
- Automatic ICP scoring
- AI-personalized email sequences
VM On-Premise
In your infrastructure
- Pre-configured VM image
- Docker Compose — all services isolated
- Compatible with VMware, Proxmox, Hyper-V
- Local AI possible (DeepSeek, Qwen...)
- 100% auditable traffic
- Secure updates
MCP connector
Per connector + 200 EUR/month maintenance
- Custom development
- SIEM, GLPI, AD, accounting, ATS...
- Proxy Aggregator NextGen (Zero-Trust)
- Smart component startup — 0 MB RAM at rest
- Pre-Hook: mandatory human validation
- Context Pruning — controlled API costs
- Complete audit logs
Price simulator
Check the modules you need and see the price in real time.
Choose your plan
Each plan includes an "AI Ready" implementation service (one-shot) followed by a monthly license for support, updates and access to the agents.
Starter
Installation + configuration of 1 Forge
Support, updates, agent access
- Audit of your IT system and needs
- 1 Forge installed and configured
- Training for your team
- Go-live with hands-on support
- Standard support included in the license
Pro
Installation + all Forges + training
Priority support, updates, agent access
- In-depth audit of your IT system and workflows
- All Forges installed and configured
- Complete training for your team
- Go-live with hands-on support
- Priority support — response within 4h
Enterprise
Multi-site, custom connectors, white-label
Dedicated support + SLA, updates, agent access
- Multi-site audit of your infrastructure
- All Forges + custom MCP connectors
- White-label — your brand, your colors
- Multi-site deployment / multi-client
- Dedicated support + guaranteed SLA
| Feature | Starter | Pro | Enterprise |
|---|---|---|---|
| Access | |||
| Forges (expertise centers) | 1 of your choice | 1 of your choice | All (5) |
| Audit frameworks | Tied to the Forge | All | All |
| Skills marketplace | Yes | Yes | Yes |
| Workspace (chat, leads, journal) | Yes | Yes | Yes |
| Agent Monitor + Pre-Hook | Yes | Yes | Yes |
| Usage | |||
| Missions / conversations | Unlimited | Unlimited | Unlimited |
| Audits | Unlimited | Unlimited | Unlimited |
| LLM cost | Your subscription | Your subscription | Your subscription |
| Infrastructure | |||
| Users | 1 | 5 | Unlimited |
| Machines | 1 | 1 | 5 |
| MCP connectors | Service | Service | 1 included/year |
| White-label | — | — | Yes |
| Support | Standard | Priority | Dedicated + SLA |
Comparison with the market
What the same functional coverage would cost with separate tools.
| Need | Market alternative | Market price | Hub Forge AI |
|---|---|---|---|
| Specialized AI agents | CrewAI / Relevance AI | 200 — 10 000 EUR/month | Included in the platform |
| AI cybersecurity expertise | CrowdStrike Charlotte AI | 4 000+ EUR/month (undisclosed) | 49 EUR/month (CyberForge) |
| Audit & compliance | Vanta / Drata | 600 — 2 500 EUR/month | 79 EUR/month per segment |
| B2B prospecting + verified decision-maker email | Apollo.io / Pharow / Kaspr | 50 — 200 EUR/month/user + credits | 69 EUR/month — Golden Records included |
| ITSM + AI | ServiceNow / Zendesk AI | 160 — 300 EUR/user/month | MCP connector (service) |
| Integration / automation | Workato / n8n Enterprise | 200 — 2 000 EUR/month | Custom MCP connectors |
| On-premise deployment | No comparable alternative | — | 199 EUR/month (VM) |
Context Pruning — cost control, maximum performance
Our proprietary deterministic context compression technology guarantees controlled cloud API costs and minimal latency — with no compromise on your agents' contextual memory.
100% local and deterministic compression
The Hub automatically extracts the key metadata from each session. This extraction is purely algorithmic: zero LLM tokens consumed for the compression itself.
Near-infinite contextual memory
Each pruning cycle preserves recent messages and keeps context between compressions. The result: a SOC agent can analyze thousands of SIEM alerts over sessions lasting several hours without ever losing the thread or blowing up the context window.
API costs divided by 3 to 8x
Without pruning, a long audit session sends the entire history with every request — linear cost explosion. Our pruner keeps a fixed and predictable ceiling regardless of session duration. The tokens saved are a direct saving on your API bill.
Frequently asked questions
Do I need an AI to use Hub Forge AI?
No, not for every module. The audit frameworks (scoring, questionnaire, gaps) and LeadForge (search, CRM, pipeline) work without AI. AI is required for the expertise centers (agents), the MCP connectors, and the advanced functions (auto quotes, reports, email generation).
Which AI is compatible?
All of them. Hub Forge AI plugs into your existing tool (Claude, Copilot, Cursor), into a cloud API (Claude API, OpenAI, Gemini, Mistral), or into a 100% local AI (DeepSeek, Qwen, Llama via Ollama, vLLM, LM Studio...). No lock-in.
Are the MCP connectors included in the plans?
MCP connectors are custom services developed for your environment. They are billed separately (1 500 to 5 000 EUR per connector depending on complexity) with optional maintenance at 200 EUR/month. The Enterprise plan includes 1 connector in the first year.
Can I change plans along the way?
Yes. You can add or remove modules at any time. Switching from one plan to another is possible mid-contract — the difference is adjusted on a pro rata basis.
What if I want to resell under my own brand (white-label)?
White-label is available on the Enterprise plan. Your brand, your colors, your templates. The Hub Forge AI engine stays invisible to your clients. Contact us for the commercial terms.
How much does the on-premise VM deployment cost?
The on-premise VM module is 199 EUR/month and includes the VM image, updates and deployment support. Thanks to smart component startup, a standard VM is enough to run the entire Hub with Ollama. The initial installation in your infrastructure is a one-time service whose cost depends on your environment (contact us for a quote).
How does Context Pruning work to control costs?
Our Context Pruning is a deterministic algorithm that compresses session history without ever calling the LLM. It automatically extracts the key metadata and injects a compact summary at the start of the window. The result: your agents can work over sessions lasting several hours (massive SIEM analysis, full audit) while staying under a fixed and predictable ceiling. According to our measurements, this divides cloud API costs by 3 to 8x on long sessions.
Book your deployment
Every company is different. Contact us for a 30-minute chat — we audit your needs and send you a quote within 48h.