Versely Blog
Page 33 of 119.
Reference-to-video is files, not adjectives. Six published rows: Veo 3.1 R2V (8s, 128cr, native audio), Happy Horse 1.0 R2V (3–15s, 34cr), Seedance 2.0 Fast R2V (4–15s, 55cr, native audio), Wan 2.7 R2V (2–10s, 20cr), Kling O3 Standard R2V (3–15s, 45cr, native audio), Gemini Omni Video (4–10s, 29cr, also R2V).
Most published video rows cap at 15s. These seven do not: Seedance 2.5 (30s, 47cr, 720p), Grok Imagine Video (30s, 7cr, 720p), Sora 2 T2V (20s, 10cr, 720p — API sunsets 2026-09-24), Sora 2 Pro Storyboard (25s, 30cr), LTXV2 Fast (20s, 4cr, 4K), Flux 3 T2V (20s, 9cr), PrunaAI P-Video (20s, 2cr, 1080p).
Type in the still is a generate of letters, not a caption overlay. Eight published still rows that actually letter: Ideogram V4 (6cr), Seedream 5 Pro (7cr), Recraft V4 (5cr), Recraft 4.1 (4cr), Qwen Image 3 (4cr / 2K), Qwen Image (2cr / 4K), GPT Image 2 (4cr / 4K), GPT Image 1.5 (2cr / 4K).
Qwen Image Edit is 4K at 3cr discounted: put the slope percent on the photographed run; /for/home-builders is the house, this is the ramp SKU, and a generated wheelchair user is not an extra.
The Chart.yaml / values.yaml listing hero letters in Recraft V4 at 5cr; not a talking SRE or a cluster demo.
As of mid-August 2026, Chinese developers occupy nine of the top ten text-to-video rows on Artificial Analysis; Google is the exception. An Elo table is not a job table — pick Seedance 2.5, Kling 3 Turbo, Happy Horse, Wan, Vidu or Hailuo by the published row.
Name a focal length and what is sharp. "Cinematic bokeh" is a vibe; "85mm, product in focus, background two stops down" is a frame. Pair it with the model that actually holds the still — Seedance 2.5 I2V via the image-to-video tool, not a restage.
Disc-golf baskets already spent the plate-and-no-throwers job: this leftover is the glow disc SKU, overlay pigment and charge time as authored type, Grok Imagine Image is 4K at 2cr for the catalog still, and do not generate throwers.
Ideogram V4 is 6cr and letters well, but Grade 2 cells on the sign you mill are raised beads a generator will invent; photograph the plaque, overlay the ADA string as type, and leave /for/sign-companies as the shop page, not this tactile SKU.
Three different 30-second stories: Grok Imagine Video is a 30s menu at 7cr / 720p with native audio; Seedance 2.5 is a 30s single pass at 47cr / 720p with native audio; Veo 3.1 extend is a 7s 720p continuation at 40cr, not a 30s generate.
Grok Imagine Video on Versely is 6 / 10 / 15 / 20 / 30s, 720p, 7 credits, native audio. The X composer is how xAI ships Imagine to a feed. It is not captions, not Seedance, not a 4K Veo take.
Hailuo 2.3 Fast is 6s or 10s at 28 credits discounted. Hailuo 2.3 Standard is the same durations at 28 credits with audio false. MiniMax H3 is 2K, 5–15s, 26 credits. Draft on Fast; keep on H3 or Veo. The MiniMax app is not the rest of the catalog.
Happy Horse 1.0 led blind no-audio arenas in 2026; the Versely row is `/models/happy-horse-1-0-text-to-video` at 1080p, 3–15s, 28 credits. An Elo win is not a studio, and it is not Kling 3 Turbo at 12 credits with audio false.
A storyboard is cells, not a paragraph. One cell, one generate, one duration the catalog actually sells — then stitch. Asking Veo 3.1 for five setups in eight seconds is how the board gets thrown away.
Seedance 2.5 on Versely is 4–30s, 720p, 47 credits, native audio. Write the thirty seconds as one continuous action, not five ads glued in a prompt. 4K is a different row.
Seedance 2.5 will hold 30 seconds at 720p if the brief is one shot with tempo shifts, not a hidden movie. Timestamp the beats; do not recast at 0:12.
Veo 3.1 is 4 / 6 / 8s, 4K, 40 credits, native audio. A 30-second product demo is Seedance 2.5 at 47 credits / 720p, or Grok Imagine Video at 7 credits with a 30s menu — not a longer Veo prompt.
Photograph the real hand and the real SKU, then I2V on Seedance 2.5 via `/tools/ai-image-to-video` (4–30s, 720p, 47cr, native audio). A generated creator holding a cousin bottle is a different identity you do not own.
imagine-heart-1-5 is a text-to-image row at 4 credits and 4K — the same credit figure as Imagen 4, a different provider page, not a video family.
Instagram did not ban AI Reels. Discovery drops when borders, watermarks, speed changes, or credit-only edits dominate. A directed generate can be original.
Leonardo Lucid Origin on Versely is a 4K still at 4 credits. Motion is a second job — `/tools/ai-image-to-video` plus a video row. Staying in Leonardo for captions, lipsync and 30-second takes is how a stills app becomes a fake studio.
LinkedIn added a report path for AI slop comments. That is comment spam, not a ban on founder video. A B2B reel still has to be a shot you directed — Seedance I2V from a real still, not a generated boardroom.
Unguarded eave in A, the same fascia with the micro-mesh hood seated in B: Veo first-last is 4s / 6s / 8s, 4K, 40cr, native audio as rain not a talking installer; leaf-guard OEM already spent the mesh SKU, and this leftover is the install on this eave.
Ideogram V4 (5 credits) letters the mark. Imagen 4 is retired on Versely. Nano Banana 2 (4 credits) edits an approved still. Overlay keeps type editable.
