Enterprise
Your models. Your GPUs. Your keys. Ship faster without your code, prompts, or IP ever leaving your control.
developers worldwide
VS Code installs
critical incident response
Trusted across regulated industries
Reserved hardware, your weights
Your models run on hardware reserved for you — not shared inference pools. Pin the region, scale capacity, and run frontier, open-source, or your own weights without noisy neighbors or rate-limit contention.
Single-tenant execution
Every agent run, code execution, and tool call happens in single-tenant sandboxes provisioned per workspace. No cross-customer process, network, or filesystem access — ever.
Zero-knowledge architecture
Encrypted from your IDE to inference and back. Customer-managed keys, per-tenant encryption, and encrypted memory mean not even Blackbox can read your prompts, completions, or code.
Configure every layer
Bring your own models, IdP, git, observability stack, DLP rules, and deployment target. Every policy and integration is configurable per workspace to fit your security model and your stack.
Nothing logged, nothing trained
Prompts and completions are discarded after every response. Abuse monitoring off, no human review, no side-channel collection. Your code never trains a model — yours or anyone else's.
Your cloud, your datacenter
Run on our hardened multi-tenant SaaS, a dedicated VPC in AWS/Azure/GCP, on-premise inside your datacenter, or fully air-gapped with no outbound internet. Regional residency in US or EU on every mode.
Tier 01 · Shared
The fastest path to production — our hardened cloud with SSO, SCIM, RBAC, audit logs, and Zero Data Retention enabled by default on Pro and above.
| Capability | What You Get | Included In |
|---|---|---|
| Security & Compliance | ||
| Zero Data Retention | Prompts and completions discarded after every response | Pro & above |
| Abuse Monitoring OFF | No human review, no prompt logging, no side-channel collection | Enterprise ZDR |
| End-to-End Encryption | Zero-knowledge chat — not even Blackbox can read your content | Enterprise |
| SOC 2 Type II | Audit in progress — interim letter available under NDA | In progress |
| ISO 27001 | Certification in progress — current scope available under NDA | In progress |
| HIPAA BAA | Business Associate Agreements for healthcare customers | Enterprise |
| GDPR & CCPA | DPA signed — data subject rights fully honored | All Plans |
| Identity & Access | ||
| SAML 2.0 SSO | Okta, Azure AD / Entra ID, Google Workspace, OneLogin | Enterprise |
| SCIM 2.0 Provisioning | Automated user and group lifecycle from your IdP | Enterprise |
| Fine-grained RBAC | Control access to repos, models, and agent capabilities per role | Enterprise |
| Audit Logs → SIEM | Streaming to Splunk, Datadog, Sumo Logic, Elastic via webhook or S3 | Enterprise |
| Session & Device Policy | IP allow-listing, session timeouts, enforced MFA | Enterprise |
| Deployment & Data Residency | ||
| Dedicated GPU Pool | Reserved GPU capacity for your models — no shared inference, no rate-limit contention | Enterprise |
| Single-tenant Sandboxes | Per-workspace isolated execution for agents, code runs, and tools | Enterprise |
| Multi-tenant SaaS | Fastest path to production on our hardened cloud | All Plans |
| Dedicated VPC | Single-tenant cloud deployment in AWS / Azure / GCP | Enterprise |
| On-Premise | Deploy inside your own datacenter with our install package | Enterprise |
| Air-gapped | No outbound internet required — fully sovereign install | Enterprise |
| Regional Residency | Pin to US or EU regions — data never leaves the region | Enterprise |
| Customer-managed Keys | BYOK / CMK for full cryptographic control | Enterprise |
| Models & Integrations | ||
| Frontier + OSS Models | Claude, GPT, Gemini, Grok, Llama, Mistral, DeepSeek, Qwen — one key, every model, automatic failover | All Plans |
| BYO Model Endpoints | Route to private models on your preferred hyperscaler or any OpenAI-compatible endpoint you host | Enterprise |
| Self-hosted Git | GitHub Enterprise Server, GitLab Self-Managed, Bitbucket Data Center | Enterprise |
| Custom Integrations | Jira, Slack, Teams, Linear, and custom webhooks | Enterprise |
| Unified API | OpenAI-compatible — one key across every model | All Plans |
| Governance & Admin | ||
| Workspace Isolation | Per-tenant encryption keys, no cross-tenant access | Enterprise |
| Usage Quotas | Per-user and per-team spend limits with real-time alerts | Enterprise |
| Model Allow-listing | Restrict which models can be used per team or project | Enterprise |
| Content Policy Controls | Custom DLP rules, prompt filters, and redaction policies | Enterprise |
| Admin Dashboards | Usage, spend, and activity across every workspace | Enterprise |
| Support & SLA | ||
| 99.9% Uptime SLA | Contractual uptime guarantee with service credits | Enterprise |
| 24/7 Priority Support | Response times as fast as 4 hours for critical incidents | Enterprise |
| Dedicated CSM | A named Customer Success Manager assigned to your account | Enterprise |
| Named Slack Channel | Shared Slack or Teams channel with our engineering team | Enterprise |
| Custom Onboarding | White-glove rollout, training, and SSO / deployment setup | Enterprise |
Encryption at rest and in transit, Zero Data Retention, abuse monitoring off, audit logs, and every compliance certification we hold.
Explore SecurityMicrosoft PartnerAzure AD / Entra ID SSO, Microsoft procurement, and AI tooling for organizations already invested in the Microsoft ecosystem.
See Microsoft DetailsPricingTransparent per-model pricing across individual, team, and enterprise tiers — with volume discounts and custom contracts for large orgs.
Compare PlansReach out to enterprise@blackbox.ai with a short description of your use case, team size, and any compliance requirements. We'll scope an evaluation that fits — including trial credits, a security review, and a path from pilot to production.
Yes. We sign custom Master Service Agreements, Data Processing Addenda, and HIPAA Business Associate Agreements. A standard DPA is available for GDPR customers, and our legal team is fast — most redlines turn around in 48 hours.
Multi-tenant SaaS, dedicated VPC in AWS / Azure / GCP, on-premise inside your own datacenter, and fully air-gapped installs with no outbound internet. Regional data residency in US and EU is available on every deployment mode.
Yes. Volume discounts start at 10+ seats. Enterprise contracts include custom token pricing, annual commitments with rollover, and usage-based billing at the workspace or department level. Net-30 and Net-60 payment terms are available.
Enterprise onboarding covers SSO / SAML configuration, SCIM setup, RBAC design, repository integration, model allow-listing, and admin training. We scope the timeline with your team based on environment complexity and compliance requirements.
Yes. Our air-gapped install ships with a curated set of open-source model weights (Llama-family, Mistral, DeepSeek, Qwen) that run fully offline inside your perimeter. For online deployments, frontier models (Claude, GPT, Gemini, Grok) are routed through the customer's preferred cloud with automatic failover across providers. The air-gapped mode is used by regulated industries with strict network egress policies.
Enterprise contracts include an uptime SLA with service credits, priority support with defined response times, and a named point of contact for ongoing account management. Specific SLA targets, response windows, and review cadences are negotiated as part of the contract.
Tell us about your security and deployment requirements — we'll scope a pilot that fits.