Letterly vs Memex

Letterly turns speech into text for messages, emails, notes, meeting summaries, journal entries, social posts, and other writing. Users can dictate inside apps such as Gmail, Slack, and Notion, or upload audio and video for transcription. Recordings can run up to 90 minutes, and speech is automatically detected across more than 90 languages, including when languages change mid-sentence. Users can select ready-made rewrite styles or create their own. Letterly can record offline, and Zapier and webhooks can send notes to other apps. Its MCP connects notes with Claude, ChatGPT, Cursor, Perplexity, Gemini CLI, Codex, VS Code, Windsurf, and Claude Code. Letterly says user words are not used to train AI; its DPA specifies encryption in transit and at rest, access controls, and regular security testing. Users can delete recordings and transcriptions through the app. It is available on Android, iOS, macOS, web, and Windows. In-app PRO prices include $19.99 a month, $89.99 a year, and $249.99, paid once.

Memex is a free, open-source journal for Android and iOS that turns text, photos, and voice recordings into organized timeline cards and insights. On-device OCR, image labeling, and photo EXIF extraction process captured material, while AI agents sort records into categories such as tasks, events, metrics, people, places, galleries, articles, and transactions. The app can show patterns as charts, timelines, quotes, and narrative summaries. Records are kept in Markdown files and a local SQLite database, with export and backup or restore options. Backups can go to iCloud Drive, a custom folder, or app storage. Users configure their own model provider, with options including OpenAI, Claude, Gemini, AWS Bedrock, OpenRouter, and Ollama. AI features require that service and may incur provider charges; text, photo metadata, OCR text, and card content may be sent to the selected provider for processing. The maker says journal data is not uploaded to a server. Memex also supports biometric app lock and custom agents with triggers, prompts, JavaScript, and inter-agent workflows.