NewPortableMind PRO Navigator, offline AI + GPS in one kit.See it →
PORTABLEMIND
Current BuildFirmware 1.8.0 · NOW SHIPPING

Firmware 1.8.0, now shipping on every USB

The biggest SCOUT Vision advancement yet: seven dedicated field scan modes with structured triage cards, danger flags, and confidence ratings. Plus faster inference, smarter hardware utilization, real-time adaptation, and a deep bug-squashing pass.

Every new PortableMind USB ships with Firmware 1.8.0 preloaded. Voice, SCOUT Vision, Phone Access, the desktop-app launcher, chat history, and AES-secured import/export, all running locally on your machine.

What changed in 1.8.0

New in this build

Firmware 1.8.0 is the current shipping build: the advanced SCOUT Vision with seven field modes and structured triage cards, faster performance across chat, voice, and SCOUT, fuller hardware utilization with real-time adaptation, and a deep bug-squashing pass.

  • SCOUT Field Modes: seven dedicated scan modes, General, Plant ID, Mushroom, Bug, Tracks, First Aid, Read Text.
  • Structured SCOUT triage cards: verdict banner, key markers, dangerous lookalikes, confidence rating, field notes.
  • Fine-tuned vision stack with danger-first safety guardrails on every identification.
  • Faster performance across chat, voice, and SCOUT, snappier model loads and switching.
  • Better hardware utilization: fuller use of available VRAM and CPU threads.
  • Smart adaptation: routing, context size, and workload placement adjust to your machine in real time.
  • Deep bug-squashing pass across the whole runtime.
Desktop experience

Launches as a desktop app

PortableMind opens as a dedicated desktop application on both macOS and Windows 10/11. No browser tabs, no terminal commands, just double-click and you're in. Voice, vision, and chat are all accessible from one clean window.

What's included

Firmware 1.8.0 feature set

Firmware 1.8.0 ships a massively advanced SCOUT Vision: seven field scan modes (General, Plant ID, Mushroom, Bug, Tracks, First Aid, Read Text), each returning a structured triage card with key markers, dangerous lookalikes, confidence ratings, and safety-first verdicts. The whole runtime also got faster, uses your hardware more fully, and adapts to your machine in real time. All local, all private. No cloud. No telemetry. Everything runs on your machine.

The biggest SCOUT upgrade yet

SCOUT Field Modes

SCOUT is no longer one generic vision box. Pick a dedicated scan mode, General, Plant ID, Mushroom, Bug, Tracks, First Aid, or Read Text, and get an answer structured for that exact job, entirely offline.

Seven scan modes: General, Plant ID, Mushroom, Bug, Tracks, First Aid, Read Text.
Structured field cards: verdict banner, key markers, dangerous lookalikes, edible parts + prep, and field notes.
Optional context input (gill color, spore print, where growing) sharpens every identification.
Sharper IDs, honest confidence

Fine-Tuned Vision

The vision stack behind SCOUT got a major fine-tuning pass. Identifications are sharper and every scan states its confidence plainly, with an explicit "if this ID is wrong" risk line so you always know the stakes.

Confidence rating on every identification, stated up front, never buried.
Danger-first triage: hazardous subjects lead with a clear DO NOT EAT / caution verdict.
Safety guardrails baked in: a photo is never treated as a green light to eat.
Quicker everywhere

Faster Performance

A performance pass across the whole runtime. Chat, voice, and SCOUT all respond faster, model loads and switches are snappier, and the app feels lighter on the same hardware you already own.

Faster inference across chat, voice, and SCOUT scans.
Snappier model loading and switching.
Lower overhead between turns, less waiting, more answering.
Your hardware, fully used

Better Utilization + Smart Adaptation

The runtime now makes fuller use of the VRAM and CPU you actually have, and continuously adapts model routing, context sizing, and workload placement to your machine and what you're doing, no config needed.

Fuller use of available VRAM and CPU threads, less idle headroom.
Routing, context size, and workload placement adapt in real time.
Works automatically on everything from tight 8 GB laptops to big GPUs.
Quieter, steadier, harder

Bug-Squashing Pass

A deep stability sweep across the codebase. Crash fixes, leak fixes, and cleaner error recovery in chat, voice, and SCOUT, so the runtime shrugs off the edge cases that used to trip it up.

Stability fixes across chat, voice, and SCOUT pipelines.
Cleaner recovery from transient errors, fewer visible failures.
Builds on the 1.7.1 hardening line, nothing regressed.
Still on the drive

Everything from 1.7

Model Manager, the Qwen 3.5 roster, multi-modal chat, GPU-accelerated voice, Phone Access, thread history, AES-hardened import/export, and the offline Field Manual all remain intact. Nothing was removed to make room.

Model Manager, Qwen 3.5 roster, multi-modal chat, GPU voice.
Phone Access, thread history, AES-256 import/export, Offline Field Manual.
Cross-platform: Windows 10/11 and macOS (Intel + Apple Silicon).
// Live UI

Eight screens from the running app

These are stills from the firmware 1.8.0 desktop app, actual frames, not mockups. Click any tile to view full-size.

PortableMind 1.8.0 desktop chat with multi-modal image attach and model pickerChat

Full desktop app, tabs for chat, voice, SCOUT. Drop an image into any chat turn and the model reads it, no mode switch required.

PortableMind 1.8.0 voice mode with GPU-accelerated Kokoro TTSVoice

Push-to-talk voice mode with adjustable transcription speed and reply style. Kokoro TTS runs on GPU via DirectML on Windows for fast, fully offline replies.

PortableMind 1.8.0 SCOUT Mushroom mode returning a DO NOT EAT verdict for an Amanita with lookalikes and risk lineSCOUT · Mushroom

SCOUT 1.8.0 in Mushroom mode. A white Amanita comes back DO NOT EAT with key markers, deadly lookalikes, and the exact stakes if the ID is wrong. All from the model on your USB.

PortableMind 1.8.0 SCOUT Bug mode identifying a regal jumping spider as harmless with confidence ratingSCOUT · Bug

Bug mode calls a regal jumping spider HARMLESS with high confidence: distinctive markers, threat profile, field notes. Seven field modes, one scan button.

PortableMind 1.8.0 SCOUT Plant ID mode flagging stinging nettle with CAUTION, edible prep, and lookalike warningsSCOUT · Plant ID

Plant ID flags stinging nettle with CAUTION: edible after thorough cooking, with lookalikes and an if-this-ID-is-wrong line. A photo is never a green light to eat.

PortableMind 1.8.0 controls panel with model picker, Model Manager entry, and field manual gridModel Manager

In-app Model Manager lists installed models, disk usage, and lets you pull any Ollama-registry model with a live progress bar. Delete gated by confirm.

PortableMind 1.8.0 phone access, QR code, local hotspot pairingPhone Access

Use your phone as a second screen for PortableMind. PC hosts a hotspot, phone scans the QR, still air-gapped from the public internet.

PortableMind 1.8.0 conversation thread with faster inference on Qwen 3.5Faster performance

1.8.0 runs the whole stack faster and uses your VRAM and CPU more fully, adapting model routing and context sizing to your machine in real time. Every answer is still 100% local.

// Hardening

Built to stay stable offline

Deep stability sweep across chat, voice, and SCOUT: crash fixes, leak fixes, cleaner error recovery.
SCOUT scan pipeline hardened for the new field modes, structured results render reliably on every scan.
Faster inference paths with tighter resource cleanup between turns.
Smarter VRAM and CPU utilization, the runtime uses more of what your machine actually has.
Adaptive routing refinements: model choice and context sizing track your hardware in real time.
All 1.7.1 hardening (TTS mutex coverage, stream timeouts, EOF retries) carried forward intact.
Why it matters

Hardening in 1.8.0 is a deep stability sweep: crash fixes, leak fixes, and cleaner error recovery across the chat, voice, and SCOUT pipelines, on top of the 1.7.1 hardening line. File custody stays intact with AES protection on packs. Offline stays the default.

How you get it

New customers

New units ship with 1.8.0

Every PortableMind USB leaving the shop is preloaded with Firmware 1.8.0, voice, SCOUT Vision, phone access, the desktop-app launcher, and all current presets. Plug in and you're on the latest build.

Existing owners

Optional update packs

Existing owners can apply an update pack when ready. Your previous firmware keeps running offline until you switch. No pressure, no forced updates.

Offline-first, with no telemetry

No cloud. No telemetry. Everything runs on your machine. Updates are optional, data stays on your drive, and the tools stay usable when everything around you is offline.