Product / Android · Development alpha
Carry the same nine-provider Valence stack — cloud, enterprise, and on-device local — into voice, widgets, quick settings, and local memory without adding a Helix middleman between you and the models.
Mobile core
The Android version keeps the cross-model core intact while adapting the working surface to pockets, commutes, quick capture, and short bursts of attention.
Continuity
Start reasoning with Claude, get a second opinion from GPT-4o, summarize with Gemini, and keep the same thread moving from your phone.
@mentions keep rerouting fast on mobileMemory
Conversations are indexed into local memory on the handset so #recall and #remember stay close to the work instead of depending on a remote proxy.
Grounding
Gemini can lean on real-time web search when freshness matters, returning cited results instead of pretending static knowledge is enough.
Input
Use speech input for hands-free prompting when typing is friction. Voice works with the same provider choice instead of locking you to a single assistant stack.
Attachments
Attach photos or files directly to prompts for models that can reason over images and mixed content, which is especially useful on a device with a camera always in reach.
Trust boundary
API keys are encrypted with AES-256 through Android Keystore and sent directly to providers. Conversations stay on the device unless you choose a remote model.
Provider ecosystem
Valence Mobile keeps the same model menu and local-first posture as the desktop product, so moving between devices does not mean downgrading the workflow.
Useful for fast generalist answers, multimodal prompts, and broad app utility.
Cloud
Strong reasoning and careful review when the mobile prompt still needs depth.
Cloud
Pairs well with search-grounded mobile questions and live information requests.
Cloud
Fits organizations that need Azure-hosted deployments and tighter platform controls.
Enterprise
Adds another response style when multiple viewpoints improve the decision.
Cloud
Helpful when the mobile workflow benefits from broader model access under one provider umbrella.
Gateway
The private path when the phone talks to local infrastructure instead of public APIs.
Local
Android native
Valence Mobile is built with .NET MAUI and uses Android-native APIs where they reduce friction: launch surfaces, quick actions, background durability, and assistant access.
Widget
Keep a Valence entry point on the home screen for quick chat starts, voice capture, and at-a-glance continuity.
Quick access
Drop straight into Valence from the Android notification shade without hunting through the launcher.
Assistant
Long-press home and launch directly into Valence so voice entry can target any provider, not just a single bundled model.
Continuity
Keep responses close while staying in another app, then expand the thread when you need to re-enter the full workspace.
Launch paths
Expose quick actions from the app icon for new chat, current thread return, and fast voice capture.
Durability
Conversations and memory persist when Android backgrounds the app, which matters on devices aggressive about lifecycle cleanup.
Mobile workflows
A few moments where the Android version changes how the product fits into daily work: quick capture, voice access, continuity, and privacy on the move.
Ask current questions while commuting or between meetings, then switch providers when one answer needs a deeper second pass.
When typing is the bottleneck, capture prompts by voice and keep the provider choice open instead of defaulting to a single assistant stack.
Use #remember when an idea lands, then pull it back later with #recall from phone or desktop.
Keep the same providers, profiles, and logic on both platforms so leaving the desk does not reset the system.
Route quick tasks to cheaper models, reserve premium models for harder questions, and lean on Ollama when local infrastructure is available.
Local storage, Android Keystore, and direct provider calls keep the trust boundary readable even on a device you carry everywhere.
Next step
Valence Mobile extends the Helix product posture to Android: provider choice, local control, native touchpoints, and direct calls to the models you trust.