1:08
Bricks.ai - Accélérez vos ventes
Bricks.AI · Oct 2019
Explore
AirCaps AI - Introducing A5S Async V1: the world's most accurate speech recognitio
AirCaps AI · Nov 2026 · 1:18 · AI-annotated
- Length
- 1:18
- Scenes
- 28 · 2.8 s avg
- Product on screen
- 0:00
- Call to action
- 1:14 · 95%
- Voiceover
- Yes
00:00:00 / 01:18:21
AirCaps AI’s 1:18 launch film runs 28 scenes, about 2.8 s each. The product is on screen from the first frame, with social proof. The call to action lands at 1:14 (95% in), over a voiceover. Locked-off near-black frames carry serif copy that blurs in and snaps sharp, interleaved with data cards that count up and take a light sweep across the top edge. The signature moves are the blur-to-focus text reveal, the rings expanding around the waveform, and the side-by-side transcript panels that type out line by line.
No filming indicatedEstimated needs: 1 capture group — your screens or recordings · 1 official logo — reused across the film · 6 graphic passages — draw or supply
Is this your launch? to your site or press kit.
More like this
0:42
Shopify Editions Winter ’23 is live
Shopify · Business software · Feb 2023
1:07
Introducing Orgo - Computers for AI Agents
Orgo · B2B · Nov 2025
AirCaps AI - Introducing A5S Async V1: the world's most accurate speech recognitio, scene by scene
· 28 scenes as text
01 · Product intro · 0:00–0:00
A centered white serif wordmark fades in on a black background, reading AirCaps with a small circular icon, then settles into © AirCaps AI.
On-screen text: “© AirCaps AI”
Camera: Static
02 · Product intro · 0:00–0:05
On a black field, the white serif lockup © AirCaps AI holds centered, then a thin vertical bar and the version string A5S Async v1 slide in to its right. The line fades out as a white multi-line audio waveform, ringed by thin expanding circles, fades in at center.
On-screen text: “© AirCaps AI”, “A5S Async v1”
Techniques: Kinetic type
Camera: Pull out
03 · Social proof · 0:05–0:07
A dark near-black frame holds a horizontal multi-line white waveform that oscillates at the center while thin white concentric rings expand outward from the middle. A faint white word Number fades in over the waveform near the end.
On-screen text: “Number”
Camera: Push in
04 · Social proof · 0:07–0:11
On a black background, white serif text enters word by word to form “Number 1 on Voice Arena's diarization bench”, then the line is replaced by a large gray serif “by a mile” that holds still.
On-screen text: “Number 1 on Voice Arena's diarization bench”, “by a mile”
Techniques: Kinetic type
Camera: Static
05 · Social proof · 0:11–0:16
A large dark serif headline, “by a mile,” fades out on a black background. Five rounded bars rise from the bottom edge, the leftmost purple and the rest gray, in ascending height, each topped by a two-line white label naming a transcription model.
On-screen text: “by a mile”, “AIRCAPS AI A5S Async v1”, “META Transcribe”, “ASSEMBLY YAI Universal-3.5 Pro”, “ELEVENLABS Scribe V2”, “NVIDIA Nemotron 3.5 Transcribe”
Camera: Pull out
06 · Social proof · 0:16–0:16
The ElevenLabs wordmark sits centered on a near-black field, rendered in dim gray and softly out of focus as it begins to resolve into view.
Camera: Static
07 · Social proof · 0:16–0:18
On a black background, white brand names slide in at center: ElevenLabs, then AssemblyAI with its angular A mark. Each name holds briefly before the next enters, matching the voiceover list.
On-screen text: “ElevenLabs”, “AssemblyAI”
Techniques: Kinetic type
Camera: Static
08 · Social proof · 0:18–0:21
White brand wordmarks appear one at a time, centered on a near-black frame with a faint warm gradient. The AssemblyAI mark slides in and holds, then gives way to Deepgram, then Microsoft, each locked off as the voiceover names them.
On-screen text: “AssemblyAI”, “Deepgram”, “Microsoft”
Techniques: Kinetic type
Camera: Static
09 · Social proof · 0:21–0:22
On a near-black background, three partner marks appear in sequence. The Microsoft window logo and wordmark slide in, then the Meta wordmark slides in, then the NVIDIA eye mark and wordmark scale up to the center.
On-screen text: “Microsoft”, “Meta”, “nVIDIA.”
Camera: Static
10 · Social proof · 0:22–0:23
The white NVIDIA eye symbol and wordmark sit centered on a near-black background. The lockup scales down slightly with a touch of motion blur while the frame stays locked off.
Camera: Static
11 · Social proof · 0:23–0:25
A dark row of benchmark bar-chart cards (AMI, AISHELL-4, CHIME-6) scrolls slowly rightward across the frame, each card listing error rates for several AI services as horizontal bars.
Camera: Push in
12 · Social proof · 0:25–0:27
A dark bar chart titled AISHELL-4 opens soft and out of focus as its values count up, then snaps sharp. It cuts to a second chart, CHiME-6, where a purple AirCaps AI bar tops a list of gray bars.
On-screen text: “AISHELL-4”, “20 Mandarin meetings · 12.7 h”, “Next-best: 5% more errors”, “CHiME-6”, “2 dinner parties · 5.2 h”, “Next-best: 20% more errors”
Camera: Static
13 · Social proof · 0:27–0:28
A horizontal bar chart on a black background compares the error rates of nine AI models, with AirCaps AI highlighted in purple at the top. The title, subtitle and a purple caption swap between two datasets, and the gray bars resize to match the new values.
On-screen text: “NOTSOFAR-1”, “160 office meetings · 16.7 h”, “Next-best: 45% more errors”, “NOTSOFAR-2”, “180 office meetings · 19.2 h”, “Next-best: 40% more errors”
Camera: Static
14 · Social proof · 0:28–0:29
A near-black background holds a single centered line in white serif. The line starts soft and slightly small, then sharpens and scales up to full size over the first half second, then holds static for the rest of the shot.
On-screen text: “The next best model is”
Techniques: Kinetic type
Camera: Static
15 · Social proof · 0:29–0:31
A near-black frame holds the phrase “40% worse” in a light gray serif. The line opens soft and out of focus, then snaps sharp while scaling up. A smaller line, “The next best model is,” sits above it as the caption resolves.
On-screen text: “The next best model is”, “40% worse”
Techniques: Kinetic type
Camera: Static
16 · Social proof · 0:31–0:32
On a near-black background, the white serif line “See the difference for yourself” scales up and sharpens from a soft blur into the center of the frame, following a brief “40% worse” caption. The frame stays locked off.
On-screen text: “40% worse”, “See the difference for yourself”
Techniques: Kinetic type
Camera: Static
17 · Demo · 0:32–0:35
A dark audio-transcription UI sharpens out of blur, showing a waveform, colored speaker bars and model scores. Two transcript cards labeled AirCaps A5S Async v1 and Deepgram Nova-3 slide up side by side, their numbered lines filling in.
On-screen text: “See the difference for yourself”
Techniques: Kinetic type, Split screen
Camera: Pull out
18 · Demo · 0:35–0:38
Two transcription panels sit side by side over a blurred office meeting video, each filling line by line with numbered segments while their accuracy percentages climb. The camera then pushes in on the left panel.
On-screen text: “5 people in an office meeting”, “0:16 / 0:33”
Techniques: Split screen, UI zoom
Camera: Push in
19 · Demo · 0:38–0:43
Two glowing panels sit side by side on black, each listing transcribed lines with numbered color badges; words type in one by one, some struck through or marked red. The view pans right then pulls back to show both, with model names and accuracy in the headers.
Techniques: Kinetic type, Split screen
Camera: Pull out
20 · Demo · 0:43–0:47
Two transcript panels sit side by side on a dark blurred background. The left gold-bordered panel and right red-bordered panel each list numbered speaker lines that scroll upward as new lines appear, while the accuracy figures in their headers tick upward.
Techniques: Split screen
Camera: Static
21 · Other · 0:47–0:52
On a near-black background, white serif copy builds line by line: each phrase starts large and blurred, then scales down to a sharp, smaller size. A glowing white sine-wave line undulates behind the first line, then fades. The final line, “Frontier accuracy without,” enters large and blurs.
On-screen text: “We're so confident our model is the best”, “that we've open-sourced our benchmark code”, “Frontier accuracy without”
Techniques: Kinetic type
Camera: Static
22 · Other · 0:52–0:55
A dark frame holds centered white serif text. “Frontier accuracy without” scales in, then yields to “sacrificing speed or cost” sliding across the middle, while thin white waveforms drift along the lower edge.
On-screen text: “Frontier accuracy without”, “sacrificing speed or cost”
Techniques: Kinetic type
Camera: Pan
23 · Other · 0:55–0:57
A dark rounded card sits centered on a black field, a glint sweeping its top edge. It reads “1 hour Audio” while a duplicate card slides left and fades, then the card’s text switches to “9.8s to process”.
On-screen text: “1 hour”, “Audio”, “9.8s”, “to process”, “sacrificing speed or cost”
Techniques: Kinetic type, Light sweep
Camera: Static
24 · Other · 0:57–1:01
A row of dark rounded stat cards slides left across a black field. A light sweep glides over the top edge of the freshly revealed card, 18¢ per hour, as the earlier 1 hour Audio and 9.8s to process cards blur and exit left.
On-screen text: “1 hour”, “Audio”, “9.8s”, “to process”, “18¢”, “per hour”
Techniques: Light sweep
Camera: Static
25 · Other · 1:01–1:04
Black frame, locked-off. A thin light-gray serif title sits centered; its letters start condensed and overlapping, then track out into even spacing to read A5S Async V1, holding still for the rest of the beat.
On-screen text: “A5S Async V1”
Techniques: Kinetic type
Camera: Static
26 · Other · 1:04–1:11
A white serif line assembles word by word on a near-black textured background: 'A new standard for speaker attributed transcription.' Three short lines then fade in in sequence: 'Put it to the test.', 'Across 9 different languages.', 'With 5 hours free.' The frame is locked off.
On-screen text: “A5S Async V1”, “A new standard for speaker attributed transcription.”, “Put it to the test.”, “Across 9 different languages.”, “With 5 hours free.”
Techniques: Kinetic type
Camera: Static
27 · Other · 1:11–1:14
On a near-black background, the serif line “With 5 hours free.” sits centered beside a small circular mark, then swaps in place to the “AirCaps AI” wordmark. Both fade out as thin white arcs curve in from the top and bottom corners.
On-screen text: “With 5 hours free.”, “© AirCaps AI”
Camera: Static
28 · CTA · 1:14–1:18
Glowing white circular arcs contract toward the center on a black field and dissolve, while the centered offer line and the URL aircaps.ai stay still, then both fade to black.
On-screen text: “Try A5S Async V1 today | Get 5h of API usage for free”, “aircaps.ai”
Camera: Pull out