Questions before you deploy.

Alluqi is the company behind Alluqi Bridge, a private-capable agent harness for the models you choose, with controlled tools and workflows across Bridge Desktop and Bridge Mobile.

Bridge Desktop is the Windows and macOS application where the agent, models, tools, and active work run. Bridge Mobile is the paired phone surface that securely continues an active Bridge Desktop session; neither surface determines where inference runs.

Alluqi Bridge is in private early access. There is no public download, App Store or Play Store release, guaranteed admission, or published release date.

It is for organizations with a defined workflow or deployment need that can work directly with Alluqi during configuration and feedback. Fit and admission are assessed individually; submitting a request does not guarantee admission.

Bridge Desktop supports Windows and macOS in private early access. Bridge Mobile platform and installation details are confirmed during onboarding; no public mobile-store release is promised.

No. Alluqi Bridge can support technical and operational workflows when the selected models, tools, permissions, and deployment are configured for that work.

Alluqi Bridge can be configured for compatible cloud, private, or local model endpoints. The specific model, provider, authentication, and feature fit are confirmed during assessment rather than assumed from a broad compatibility claim.

Alluqi Bridge is the agent harness, not the underlying model. A deployment connects models selected for the workload; provider access, model licensing, and private hardware are scoped separately where needed.

Yes. Private or local inference always needs a physical GPU-capable workstation or server somewhere. That hardware can be in the customer environment or provided as a dedicated Alluqi Managed server.

Possibly. Alluqi evaluates the workstation’s GPU memory, system memory, operating constraints, target model, context needs, and expected workload before confirming that it is a practical private-inference host.

A dedicated server may be appropriate when the selected model, concurrent users, workload duration, availability needs, or operating controls exceed what an existing workstation can support. The decision follows an assessment, not a fixed threshold.

Yes, when the endpoint, authentication, network path, and model interface are compatible. Alluqi validates access and security boundaries during configuration before connecting tools or production workflows.

Model connections can be changed when compatible alternatives are configured, while the surrounding workflow can remain familiar. Output quality, context, tool support, and provider behavior may still require workflow adjustments.

Sizing considers model size and format, context length, concurrent use, latency expectations, workload duration, integrations, and operating headroom. Alluqi recommends capacity after testing the actual use case rather than promising a universal configuration.

Cloud-only uses selected cloud model providers and requires no dedicated private-inference server. Private-only uses a physical GPU-capable workstation or server and does not send inference to a cloud model provider. Hybrid configures both and lets the workload determine which model is used.

Hybrid means both private or local models and cloud models are configured in the same operating approach. It does not mean Bridge Mobile access, and it does not describe who owns the hardware.

On-prem means the physical inference hardware is installed in the customer’s environment or another location the customer controls. An Alluqi Managed server is hosted outside that environment, so it is not on-prem.

Alluqi Managed is dedicated, single-tenant, isolated rented infrastructure hosted, maintained, and operated by Alluqi under a custom monthly agreement with no public rate. It is not on-prem, and traffic must leave the customer network to reach it.

No. The server is dedicated to one customer and isolated from other customers. Its capacity, controls, access, and operating boundaries are defined in that customer’s agreement.

Alluqi provides, hosts, maintains, and operates the rented infrastructure. Equipment ownership, service term, maintenance, support, and end-of-service handling are defined in the custom agreement; the server is not a customer on-prem asset.

Cloud-model requests are processed by the selected provider. Private inference runs on the configured physical workstation or server; an Alluqi Managed server is outside the customer network. Hybrid can use either path. Alluqi Bridge transports encrypted session traffic but does not determine inference location.

Yes, subject to compatible model endpoints, hardware, network access, policy, and testing. Alluqi reassesses the changed path and reconfigures the affected controls rather than treating the move as an automatic switch.

Bridge Mobile lets a paired phone securely continue an active Bridge Desktop session. It is one Alluqi Bridge access surface across cloud-only, private-only, and hybrid strategies; it does not move the model or choose where inference runs.

Alluqi Bridge traffic is end-to-end encrypted between Bridge Mobile and Bridge Desktop. The relay carries encrypted session traffic and does not need the plaintext prompt, code, or response to route it.

No. The relay cannot read prompts, code, or responses because that session content is end-to-end encrypted between paired endpoints. This transport boundary does not hide cloud-model requests from a cloud provider you choose to use.

Yes. Bridge Mobile access to Bridge Desktop requires internet connectivity so encrypted traffic can cross the relay. Bridge Desktop and the selected model endpoint may have additional connectivity requirements based on the deployment.

Bridge Mobile is designed to reconnect a paired device after connectivity returns and resume access to the active session. Behavior for work that was in flight during a disconnect is confirmed during early-access onboarding rather than covered by a timing guarantee.

Pairing is device-specific. Supported device counts and the multi-device policy are confirmed during early-access onboarding; Alluqi does not publish an unsupported device limit.

Yes. A paired phone or computer can be revoked so it no longer has relay access. The available administration path and responsibility for revocation are confirmed during early-access onboarding.

No public App Store or Play Store release is available. Bridge Mobile is part of Alluqi Bridge private early access, with platform and installation details confirmed during onboarding; there is no promised public release date.

Data goes to a cloud model provider only when a cloud model is selected and used for that work. Private-only inference is processed on the configured private hardware and is not sent to a cloud model provider for inference.

The selected cloud model provider processes the request under that provider’s service and data-handling terms. Alluqi Bridge encryption protects relay transport, but it does not make processing private from the cloud provider receiving the request.

Storage depends on the selected deployment and configuration. Exact session, settings, retention, access, backup, and deletion boundaries are confirmed and configured during onboarding rather than implied as universal guarantees.

Permissions define which approved tools, resources, and actions a configured workflow can use. Alluqi maps those controls to the use case during onboarding; the exact boundary depends on the connected systems and customer policy.

Scope can include models and inference, hardware, tools and permissions, policy, Bridge Mobile, integrations, identity, logging, backups, operations, launch, training, and support. The exact scope is defined from the assessed workflow and environment.

Alluqi can configure suitable customer hardware or provide a dedicated Alluqi Managed rented server when private inference is required. Sizing and agreement terms follow assessment. A cloud-only strategy requires no dedicated private-inference server.

Alluqi can connect compatible systems when appropriate access, authentication, data boundaries, and APIs are available. Integration feasibility and permitted actions are validated before they are included in scope.

The timeline is scoped after an assessment of models, hardware, integrations, permissions, testing, and launch needs. Alluqi does not promise a universal delivery window before that work is understood.

Yes. Training, launch support, operational handoff, and ongoing support can be included. Coverage, responsibilities, response expectations, maintenance, backups, and availability are defined in the engagement or managed-service agreement.