You know those long naps where you wake up and forget who you are? Nativ doesn't.😌 Your Mac can doze off mid-recording and wake up with the recording saved. Nativ still knows your name, your style, and where that one chat is. v0.3.9. Smoother than your Sunday nap. Try it here:

Nativ v0.3.9 is out🚀🚀 Chats that sound like you, and search that finds them again. → Personalization for new chats: profile, response style, emoji, formatting → Local chat search: Command-F inside a chat, or search every saved one → Artifacts that hold: files keep their identity, images refresh, deletes reach every linked chat → Recording recovery: if the Mac sleeps mid-recording, the audio lands in Audio → Library → Clearer chat feedback: reading progress, smoother streaming, Escape declines a pending tool call → Better model info: sizes from real download bytes, installed models split from previously used → Sharper failure records: categories, app and runtime versions, cancellations told apart → Simpler API settings: the runtime handles cached-model discovery, so one toggle is gone Another step toward making Nativ feel fast, native, and reliable. Thanks to everyone who contributed🙏 Try Nativ:

Airplane mode on. Games still running. Both built with the Nativ integration, fully offline. Same integration organises your files, works on your code, and much more without anything leaving your Mac.😉 Try Nativ:

Nativ v0.3.7 is out 🔥🚀 Projects, and chats that travel: → Projects that group your chats around a local folder → Import and export chats as files → Model shelf, pin your favourites and remove in bulk → Set a draft model straight from chat → Granite Speech joins the speech models → Audio library with recordings, transcripts, and summaries → SearXNG for private web search inside chat → Smoother chat editing, replies, and titles → Token counts and time to first token → New Mac artwork Another step toward making Nativ feel fast, native, and reliable. Thanks to everyone contributing 🙏 Try Nativ:

Congrats to @Zai_org on GLM-5.3-Flash! Day 0 support in Nativ 🎉 320B params, 18B active · 1M native context · fully on-device. On an M3 Ultra (512GB) with Nativ v0.3.4, standard 4-bit MLX: ⚡ Up to 505 tok/s prefill · 32 tok/s decode 🚀 Batch 4 lifts decode to 70.9 tok/s 💾 Peak under 380GB, fully resident, no offload 📏 Benchmarked to 128K context A 320B MoE running entirely on your Mac. Try Nativ 👇

LFM2.5-2.6B now runs locally on your Mac 🚀 Great release by @LiquidAI — and Nativ supports it Day 0. Full BF16, no quantization: ⚡️ 11,000+ tok/s prefill ⚡️ 82 tok/s decode 🧠 Under 8.5GB peak memory — even at full 128K context 📈 Scales to 476 tok/s aggregate across 16 concurrent requests Try it on Nativ 👉 Github repo 📷