Instagram Pose: 4 Before/After Story Presets
Instagram Pose is an image-to-image preset tool for quick, trendy portrait variations. This before/after post runs one input portrait through four story-style…
Blog
Generative AI Blog by Wiro
Kling V3 Omni: 3 Sound-On Text-to-Video Tests (720p)
Kling V3 Omni is a text-to-video model that can generate motion and sound from a single first-frame image. In this post, I…
Video Background Music v2: 4 styles for one clip
Video Background Music v2 generates instrumental soundtracks that match a video. This prompt guide runs one clip through four styles so it…
MMAudio: 4 Video-to-Audio Before/After Tests
MMAudio generates synchronized audio for a video. This post runs four before-and-after tests and shows the resulting clips with audio. Model MMAudio…
Nemotron vs Whisper Large V3: 5 Audio Transcription Tests
Nemotron vs Whisper: two very different ASR approaches NVIDIA Nemotron-Speech-Streaming-En-0.6b targets low-latency streaming transcription (chunked audio) with punctuation and capitalization support. OpenAI…
Easy OCR: 5 Layout Tests
Easy OCR extracts text from images. The hard part is the input. This post runs five layout tests and shows the raw…
LTX-2.3: 5 Text-to-Video Tests at 1080p
LTX-2.3 can turn a plain text prompt into a short 1080p video. This review runs five tests with very different scenes. The…
Seed-V2 Mini vs Qwen3.5-27B: 5 Small Tests
Seed-V2 Mini vs Qwen3.5-27B sounds like a simple comparison. The outputs can look very different in practice. This post runs five small…
Text To Song: 6 Prompts for 10-Second Music Clips
Text To Song can turn a short caption and optional lyrics into a music clip. This post shares six copy-paste prompts and…
Trellis-2: 3 Image-to-3D Tests
Trellis-2: 3 Image-to-3D Tests Trellis-2 converts 2D images into production-ready 3D meshes (GLB) with PBR textures. This short review runs three quick…
Qwen3.5-27B: 6 Quick Tests on Reasoning, Parsing, and Code
Qwen3.5-27B: 6 Quick Tests on Reasoning, Parsing, and Code Qwen3.5-27B shows how a 27B multimodal model handles long-context reasoning and mixed tasks.…
8 Multi-Speaker Dialogue Prompts for FishAudio S2 Pro
FishAudio S2 Pro supports multi-speaker TTS in a single generation. These eight dialogue prompts show speaker switching, timing tags, and emotion control.…
FishAudio S2 Pro vs Qwen3-TTS: 6 Audio Tests
FishAudio S2 Pro vs Qwen3-TTS: six short audio tests compare clarity, timing, and prosody. Each test uses the same script so results…
Seedance Pro V1.5: Text-to-Video in 5 Vertical Tests
Seedance Pro V1.5 Uncensored generates short videos from text prompts. This review runs five vertical prompts that stress camera motion, texture detail,…
FireRed Image Edit 1.1: 6 Real Before and After Edits
FireRed Image Edit 1.1: 6 real before and after edits Image editing fails in boring ways. A face shifts. A logo melts.…
Animated Logo: 4 Before/After Preset Videos
Animated Logo turns a static logo into short product-style videos using preset scenes. In this before/after test, I used one simple logo…
8 Prompts for Vertical Video with Wan 2.6
Wan 2.6 can generate short vertical clips from text prompts. This prompt guide focuses on clean camera moves and ad-friendly shots. Each…
Seedance V1 Pro Fast: Fast Text-to-Video in 5 Tests
Seedance V1 Pro Fast is a fast text-to-video model that targets short clips with clean motion. This post runs five prompts that…
Seedance V1 Pro Fast vs Wan 2.6: 5 Prompt Video Test
Seedance V1 Pro Fast vs Wan 2.6 in 5 prompts Seedance V1 Pro Fast and Wan 2.6 both aim at the same…
FireRed Image Edit 1.1 vs FLUX.2-dev: 10 Edits Compared
FireRed Image Edit 1.1 vs FLUX.2-dev: 10 Edits Compared FireRed Image Edit 1.1 and FLUX.2-dev were compared across ten edits. The tests…
MOSS-TTSD: Dialogue TTS in 6 Tests
MOSS-TTSD turns a dialogue script into spoken conversation. This post runs 6 short tests and shares the raw audio outputs. The goal:…
DreamOmni2: 6 Multi-Image Before and After Edits
DreamOmni2 handles instruction-based image edits with multiple reference images. This post runs 6 before/after edits and shows the raw outputs. Model xiabs/dreamomni2…