Wan 2.6 Flash is silent 20-credit I2V

Wan 2.6 Image to Video Flash is 20 credits, image-to-video, audio false, 1080p, 5 / 10 / 15 seconds, image required. Wan 2.6 Image to Video is 20 credits (was 40), image-to-video, native audio true, 1080p, the same three lengths, image required. Flash is silent 20-credit I2V. The full row, at the same 20 credits for now, is the hero when the take needs a room, a line, or foley in-model. Draft on Flash. Ship on the full row when the audio flag is the job.

This is not an industrial “still is the contract” twin. The still matters on both rows — that is what I2V is — but this page is the flag: mute versus native audio, same resolution, same menu, and for now the same 20 credits. It is also not Wan 2.7. 2.7 I2V is 16 credits, 1080p, 2–15 seconds, audio null. Do not mix 2.6’s native-audio hero with 2.7’s null card. Wan 2.6 text-to-video is a different category at 40 credits with native audio, 5 / 10 / 15 seconds. This page is I2V only.

The Flash card: audio false, is the mute shortcut, 20 credits, 1080p, three lengths. The description is ultra-fast image-to-video. Speed is not a reason to skip the still. Silence is the reason to pick the row.

The full 2.6 I2V card: audio true, 20 credits (was 40), 1080p, three lengths, discounted on the sticker. Description: image-to-video with audio support. Same still slot. Different bill. Different soundtrack.

Flash is silent on purpose

audio: false is the useful fact. There is no native dialogue, no multilingual mouth, no soundtrack you forgot to request. You will not get a talking presenter because the prompt mentioned a line. You will get motion on the still, quiet.

That is the draft job. It is also a ship job when the feed will get VO later, or when the cut is mute autoplay and you will burn a caption instead of a voice. Captions for silent autoplay is the typesetting. Flash does not typeset. It moves the frame.

Do not prompt “add voiceover” and expect the flag to flip. The flag is the row. If the brief is a talking SKU, leave Flash. Use Wan 2.6 I2V at 20, or leave the family.

Twenty credits is the catalog sticker. Do not convert it to USD. Do not hunt for a mute discount: Flash and the full row both start at 20. The difference is mute versus native audio, not a coupon.

When Flash is the pick:

• You are testing whether this still can move without melting the label.
• The VO will be a desk agent, a creator, or Cartesia Sonic 3.5 laid after.
• The placement is mute-first and a burned-in line is the audio.
• You will throw the take away if the motion is wrong. Twenty credits is how you find out.

When Flash is the wrong pick:

• The hero needs in-model speech, room tone, or a pour that should sound like a pour.
• You already know the still is locked and the take is the ship file with native audio.
• You wanted text-to-video. Flash requires an image.

The full row is the native-audio hero

Wan 2.6 I2V at 20 credits is the same still, the same 5 / 10 / 15, the same 1080p, with audio on. Write whether anyone speaks. SKU orbits usually want foley and no voice. Say “no speech.” A 20-credit surprise monologue is still a wasted take. A pack that should speak should have a line in the prompt, then a caption pass that spells the SKU as the carton spells it.

Hero does not mean “always the full row.” Hero means the take that ships with the soundtrack the row can make. If you will replace the soundtrack in the edit, the native audio is wasted. Stay on Flash.

If you need 4K dialogue in 4 / 6 / 8 seconds, neither Wan 2.6 I2V row is Veo. Veo 3.1 is 64 credits, T2V, native audio, 4K. VEO First Last Frame is 64 credits, I2V, native audio, 4K, 4 / 6 / 8 seconds. Wan is 1080p and three long holds. Do not rerun Flash at 20 credits hoping it becomes 4K. The max on both 2.6 I2V cards is 1080p.

If you need 20 or 30 seconds, neither 2.6 I2V row grows. Flux 3 First Last Frame to Video is 34 credits, native audio, 5–20 seconds, two stills. Grok Imagine Video is 24 credits, T2V, native audio, 720p, up to 30 seconds. Length is a different comparison.

Same three lengths, same 1080p, different bill

supports_durations is 5s, 10s, 15s on both cards. There is no 6s, no 8s, no 2s test. A 5s generate is the draft length. 10s is the default ship length. 15s is the hold. Cutting a 15s Flash take down to 6s in an editor is fine; paying for 15s because you wanted 6s and the slider felt adjacent is not.

Quality settings go to 1080p. Aspects follow the still. Lock the crop on the image before you generate. The image-to-video launcher is the door. A text prompt with no file is a different category.

Draft versus hero in production order:

1. Approve the still at 100%. Soft type will not sharpen in video on either row.
2. Run Flash at 5s. If the label crawls, fix the still, not the prompt.
3. Run Flash at the ship length if mute is enough.
4. Promote the same still to 20-credit Wan 2.6 I2V only when you want native audio on a motion you already liked.

Do not run the full row first “to see the audio” on an unapproved pack. You will spend the hero run on a draft still. Do not run Flash and then be surprised it is quiet. You picked the quiet row.

Kling 3 Turbo Text to Video is also silent, 22 credits, T2V, 3–15s, 1080p. It is not this comparison. Turbo has no still slot. Flash has no sentence-only slot. Mute is not one row.

Wan 2.7 I2V at 16 credits with audio null is the easy mix-up. Null is not true. Null is not false. Do not claim native audio on 2.7. Do not treat Flash’s explicit false as the same object as 2.7’s null. This page stays on 2.6.

What you do after the mute take

A silent 20-credit clip is not a finished ad. Burn a caption. Lay a VO if the brief needs a voice. Crop. The AI caption generator is the typesetting. Cartesia Sonic 3.5 is 4 credits if you need TTS. Native audio on the full row is the other way to get a soundtrack. Pick one. Doing both without a plan is how you stack a VO on a VO.

Article 50 still applies to photoreal people in the still. Flash being mute does not exempt a generated face. Label at first exposure for EU audiences since 2 August 2026. The Article 50 disclosure template is the copy.

Sora 2 leftovers that were mute product pushes at 720p can land here at 1080p for 20 credits if you have the still. Sora’s API shuts 24 September 2026 (OpenAI Help Centre, September 2026). This is not a Sora essay. It is a mute I2V sticker.

If the still is the PDP pack, Flash is the mute way to see whether the live hero can move. If it can, and you need sound, run the full row. If it cannot, you still have to fix the still.

Flash is the mute draft. The full row is the talking hero. Same lengths. Same 1080p. Same 20 credits for now. The flag is the pick.

Run the mute draft in Versely.

FAQ

Does Flash ever speak if I ask nicely?

No. audio is false. Change rows. Wan 2.6 I2V at 20 credits is the native-audio sibling.

Is Flash a discount on the same file?

No. You get a silent file at the same 20 credits. The full row is the same motion menu with audio on, currently discounted from 40. Promote only when you want that flag.

Can I use Flash as the ship file?

Yes, when mute is the ship: VO later, or caption-only autoplay. Do not ship Flash to a brief that required in-model speech.