Samsung Adds AI Features to A37 and A57

What happened: Samsung launched the Galaxy A37 5G and Galaxy A57 5G for the US market, pricing them at $450 and $550 respectively and shipping them with expanded AI features under the Awesome Intelligence brand. Availability begins April 9, 2026, and Samsung markets these models as bringing many flagship software capabilities to mid-range devices.

---

Technical details: The two phones use Samsung’s Exynos silicon — Exynos 1480 in the A37 and Exynos 1580 in the A57 — and emphasize on-device AI-driven features rather than a raw specs race. Key consumer-facing features include: - Voice transcription inside the recorder app, designed for real-time text capture and review. - AI text extraction to pull text from images or screenshots for quick copy/paste and search. - Smarter multitasking tools that streamline split-screen and app switching workflows. - Camera AI improvements: Object Eraser for cleaner removals, Best Face on the A57 to composite group shots, and enhanced visual search that can identify multiple subjects within a frame.

Why the hardware matters: The use of Exynos 1480/Exynos 1580 signals Samsung is pushing more inference workloads onto mid-range SoCs, enabling low-latency, privacy-preserving use-cases without always routing data to cloud models. Samsung also advertises extended update commitments for these models, which affects device lifecycle and security for deployment scenarios.

---

Context and significance: Flagship devices have historically gated the most advanced AI features behind higher-end silicon and pricing. The Galaxy A57/A37 move continues a broader industry trend: shifting key AI functionality down the product stack so more users get native, on-device inference. For ML practitioners and mobile engineers, this matters because: - It raises the baseline of available client-side capabilities you can assume in consumer-facing apps (local transcription, OCR, basic vision tasks). - It pressures app and model teams to optimize for power-efficient on-device execution and mixed local/cloud inference patterns. - It creates competitive dynamics: OEMs must balance model complexity, quantization, and latency to deliver responsive UX on mid-tier NPUs.

This launch mirrors similar pushes from other vendors to bake AI into everyday workflows rather than keep it as a premium differentiator. For product teams, the practical takeaway is to re-evaluate where to offload services — leveraging client-side AI for latency-sensitive features and using cloud models for heavier tasks.

What to watch: Monitor how developers adapt to the expanded on-device feature set and whether Samsung publishes SDKs or tooling that expose Awesome Intelligence primitives to third-party apps. Also watch battery and thermal behavior under sustained inference workloads, and whether competitors match these features at comparable prices.

---

Sources: androidcentral.com news.samsung.com androidpolice.com