Welna 2.8 moves to OpenAI's GPT-6 and makes the connection to OpenAI more dependable. • Faster, more useful answers on GPT-6. Built on your own numbers, they point out what changed, what is worth noticing, and what to bring to your clinician. Each lab value and each medication on your list is explained in plain language. • More accurate error messages. When the usage included in your plan is spent, Welna now says exactly that, including when it resets, instead of showing a generic error. • Fewer failed answers. Welna now accepts replies it used to reject at the last step, so more runs finish on the first try. • Less background chatter. When Welna already knows it cannot reach OpenAI, for example while your plan's included usage is spent, it stops asking until something changes, instead of retrying quietly in the background.

Welna 2.7 makes the connection to OpenAI more stable. • Fewer repeated errors. When OpenAI is busy, or the usage included in your plan is spent, Welna now waits quietly instead of trying again and again in the background. • Clearer guidance. When a connection problem does happen, the message tells you what happened and what you can do next. • Your on-device features are unchanged. Daily reads and your lab timeline keep working without a connection.

Welna 2.6 polishes the 2.5 release. • A faster, steadier OpenAI connection. Answers start sooner, brief network hiccups recover on their own, and when something does go wrong the message tells you exactly what to do. • Import from Photos and Files. Lab reports and documents can now come from an existing photo, an image file, or a PDF, read on your device just like a scan. • Clinician handouts, sharper. The shared PDF is cleaner and easier to read, with the QR code carrying the secure link. • Lab imports, more reliable. Percent values, qualitative results, and repeated reports are handled properly, and imported names stay exactly as printed by the lab. • Metric explanations that read your week. "What this means" now interprets your recent pattern against your own usual range and says what it means for you. • Apple Watch and widget layouts refined so nothing is cut off. • Connecting Apple Health is smoother, and the permission prompt appears only when you choose to connect. • Fixes and polish across all six languages.

Welna 2.5: your health, in your language, everywhere you look. • Six languages. Welna now speaks English, French, German, Hindi, Simplified Chinese, and Spanish. • The newest GPT-5.6 models. Welna now runs on OpenAI's GPT-5.6 family through your own account: faster answers on everyday reads and workflows, and Welna stays compatible as OpenAI updates its models. • On your wrist and Lock Screen. A new Apple Watch app, complications, and widgets keep your daily read at a glance. • Siri and Shortcuts. Ask for your brief or start a visit prep hands-free, with a Live Activity while Welna works. • Your labs, over time. Import a lab PDF or photo and Welna builds an on-device timeline of your biomarkers, each shown against the lab's own printed range and your personal trend. • Explain Results, upgraded. Imported values are saved with the conversation, so you can keep asking. • One big thing. Every daily read now leads with the single insight that matters most. • A metabolic read. Glucose in context with your sleep, steps, and heart data, if you track it. • Cycle context. Gentle, optional context in morning reads if you log your cycle. Not a tracker. • Share with your clinician. A clean one-page PDF handout with a secure link and QR code, generated on your device. • Start instantly. See your first on-device read before connecting any account. • Clearer AI labeling, larger text support, reduced motion, and VoiceOver improvements throughout. As always: your text is redacted on your device before anything is sent, and Apple Health stays read-only.

A bigger Welna: understand your numbers, and talk to them. • Daily reads. Welna now writes short, on-device reads from your own Apple Health: a recovery read each morning, a momentum check at midday, an evening recap, and a periodic 90-day look. Computed on your device, and works offline. • What would help. Each read carries one gentle, grounded suggestion from your own numbers. Patterns many people find helpful, not medical advice. • Ask about anything. Tap a metric or a read and ask questions in plain language, grounded in your own history, baseline, and sleep. Your conversations are saved. • Explain your results, then keep asking. Paste or scan a lab or test result and Welna explains what each value generally means, next to what your own numbers usually show. • Notes. Write a line in your own words and Welna connects it to that day's real numbers, then keeps a simple record. Your raw words never leave your phone. • Make it yours. Choose an accent color, light or dark, and a dark style. • Privacy unchanged. Anything you type is redacted on your device before anything is sent, and only normalized summaries ever leave the phone. • Refinements and reliability throughout.

• Stronger on-device redaction. Follow-up questions inside a run are now redacted on your device before anything is sent. • If you connect Hugging Face to keep a private record of your runs now you can choose which of your accounts or organizations it is stored in. • Reliability and small refinements throughout.

A calm, refreshed look across every screen. Your Apple Health data now drives real, plain-language insights. Every metric, trend, and reading is your own data, never a sample. Today's read: a short, friendly summary of how your day is going, generated from your own numbers.