ImageFlow Brings Free Local Relighting to Browser

// 1h agoPRODUCT LAUNCH

ImageFlow Brings Free Local Relighting to Browser

ImageFlow is a free, Photoshop-grade browser editor with layers, masks, PSD support, and interactive Relight tools. Its on-device processing keeps files local while challenging Adobe’s subscription-and-credit model. [ImageFlow](https://imageflow.dev/) [Product Hunt](https://www.producthunt.com/products/imageflow-6)

// ANALYSIS

ImageFlow has not killed Adobe, but it demonstrates how local browser compute can turn privacy and zero pricing into serious competitive advantages.

–Combines advanced editing workflows with on-device AI selection and relighting.
–Local processing avoids uploads, cloud latency, subscriptions, and generative credits.
–Browser memory limits, PSD fidelity, RAW workflows, and photo catalog management remain major hurdles.
–It is currently closer to a Photoshop alternative; the dedicated Lightroom replacement is still roadmap territory.
–The open-source story appears incremental, with components planned for release rather than the entire editor being clearly open-source today.

// TAGS

imageflowvisionmultimodallocal-firstopen-source

DISCOVERED

1h ago

2026-10-10

PUBLISHED

1h ago

2026-10-10

RELEVANCE

7/ 10

AUTHOR

0x0SojalSec

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

RESEARCH47m ago

AIR-LLM Broadcasts Weights for Memory-Free Edge Inference

AIR-LLM broadcasts LLM weights from a central radio, then uses RF mixers on edge devices to perform matrix-vector operations without storing the model locally. Evaluations report a 4% perplexity gap on LLaMA-3.1-8B and up to 157.7× lower energy than FP16 inference.

UPDATE52m ago

Synara v1.0.2 Beta 1 surfaces subagents

Synara’s v1.0.2 Beta 1 makes subagent activity visible, letting developers track active and completed runs, navigate parent-child chats, inspect clearer tool summaries and timings, and pause or background work.

MODEL59m ago

Qwen3.8 Coder390 Tames Runaway Reasoning

This open-weight Qwen3.8-27B variant uses additional SFT and RLOO post-training to reduce 94K-token truncations and empty answers while improving coding and reasoning benchmarks. Static FP8 reaches 178/198 on GPQA, 450/500 on MMLU, and 90/100 on LCB. [Model card](https://huggingface.co/nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2)