Whisperila-1.0.4

WHISPERILA // RELEASE 1.0.4

[VERSION: 1.0.4] [ARCH: arm64 + x86_64] [STATUS: STABLE]

"Contextual depth, effortless control."

Whisperila 1.0.4 introduces native Markdown (.md) and plain-text file imports for AI instructions, an expandable multiline context editor, and targeted interface refinements for custom instructions in the Models panel.

---

01 // Context & AI Instructions Overhaul

• [IMPORT] Markdown & Text File Upload
  Load system prompts, tone instructions, meeting templates, and context files directly into memory using the new Upload .md / text file… action—or simply drag and drop any .md, .markdown, or .txt file straight onto the AI Instructions & Context section.
• [EDITOR] Expandable Multiline Instructions
  Replaced the single-line input with a fluid multiline editor (2...6 lines) engineered to handle structured Markdown, multiline guidelines, bulleted lists, and plain English with ease.
• [MANAGEMENT] Context Counter & Instant Reset
  Added a dynamic instruction count badge in the section header paired with a single-click Clear All action for immediate context resets.

---

02 // Interface Fixes & Refinements

• [FIX] AI Instructions "Add" Action & Feedback
  Resolved a macOS grouped-form focus issue where clicking Add failed to register while the text field remained active. Upgraded the control to a .borderedProminent button with instant Instruction Added toast confirmation.
• [REFINEMENT] Neutral Placeholder Guidance
  Replaced specific founder/company sample text with clean, universal onboarding copy: "Enter custom instructions or context in plain English or Markdown (e.g. style rules, role, format)…"

---

03 // Downloads & Verification

Architecture	File Package
Apple Silicon (M1–M4)	Whisperila-1.0.4-arm64.dmg
Intel Mac (x86_64)	Whisperila-1.0.4-x86_64.dmg

---

04 // Build Telemetry

Release Version : v1.0.4
Compatibility   : macOS 14 Sonoma or later (macOS 14, 15, and 26)
Context Formats : .md, .markdown, .txt (Native Drag & Drop + File Picker)
Trigger Shortcut: Option (⌥) + Space
Release Hub     : [https://github.com/buildbyaditya/Whisperila/releases](https://github.com/buildbyaditya/Whisperila/releases)

Whisperila-1.0.3

WHISPERILA // RELEASE 1.0.3

[VERSION: 1.0.3] [ARCH: arm64 + x86_64] [STATUS: STABLE]

"Native precision, uninterrupted flow."

Whisperila 1.0.3 expands our hardware footprint to every modern Mac, introduces granular control over your voice intelligence engines, and adds resilient failover architecture so your dictation never misses a beat.

---

01 // Architecture & Distribution

• [INSTALLERS] Dual-Architecture DMG Packages
  Standalone .dmg disk image installers are now built natively for both Apple Silicon (arm64) and Intel (x86_64) Macs—delivering uncompromised local performance across every machine.
• [UPDATES] Direct GitHub Release Navigation
  Selecting Check for Updates in the About window now routes directly to our Official GitHub Releases repository for instant version verification and manual binary downloads.

---

02 // Intelligence & Engine Control

• [CONTROL] Live Model Management
  The Models settings view now features real-time ACTIVE and INACTIVE telemetry badges alongside instant quick-activation toggles across all on-device and cloud dictation engines.
• [FAILOVER] Automatic Model Shifting
  Engineered for zero downtime. If an active cloud provider encounters rate limits or quota exhaustion, Whisperila automatically shifts to the best available installed local or cloud model in the background—keeping your ⌥ Space workflow seamless.

---

03 // Build Telemetry

Release Version : v1.0.3
Target Builds   : Whisperila-1.0.3-arm64.dmg (Apple Silicon)
                : Whisperila-1.0.3-x86_64.dmg (Intel Mac)
Failover Engine : Zero-drop automatic local/cloud routing
Trigger Shortcut: Option (⌥) + Space
Release Hub     : [https://github.com/buildbyaditya/Whisperila/releases](https://github.com/buildbyaditya/Whisperila/releases)

Whisperila-1.0.2

Whisperila 1.0.2 — Release Notes

[VERSION: 1.0.2] [PLATFORM: macOS] [ENGINE: CoreML + Metal]

Don't type, just speak.

Whisperila 1.0.2 brings zero-latency, on-device voice intelligence to macOS. Built natively for Apple Silicon, it turns spoken thoughts into clean, punctuated writing across every Mac app—with zero cloud compromise.

[100% LOCAL] Runs entirely on-device via Apple Silicon CoreML and Metal. Your voice and transcripts never leave your Mac.

[ZERO FRICTION] Hold or tap ⌥ Space anywhere to speak, and watch formatted text appear instantly at your cursor.

[EDITORIAL PRECISION] Automatic punctuation, capitalization, and smart phrasing that turns natural speech into polished prose.

[NATIVE DESIGN] Tactile macOS interface featuring fluid spring physics and the responsive 4-pillar audio waveform HUD.

What's Improved in 1.0.2

[PERFORMANCE] Faster CoreML neural engine warm-up after system wake.

[RELIABILITY] Improved text-injection stability across Electron and native macOS text fields.

[UI] Refined cubic-bezier spring transitions on the floating listening pill.

WHISPERILA — Offline speed, zero cloud compromise.