Introduction
TemPolor v4.8, MiniMax Music-3.0, Google's Lyria 3.5 and Mureka V9.5 — four AI music models, the same six briefs, one useapi.net API token. All 24 tracks are below, complete and unedited, each with the API call that made it. They are the same briefs we used in July, so every track here has a predecessor to compare against in AI Music APIs Compared.
TemPolor is the newcomer, and it prices differently from everything else here: a flat $30 a month for unlimited generations, where the other three charge about a cent a track. What that does to the maths is below. Listen first and judge the models yourself.
Short on time? There is also a 12-minute YouTube video — each track's loudest 30 seconds back to back, loudness-matched, with the prompt, price, and generation time on screen. The full-length tracks and the exact API call behind each one live on this page.
What changed since July
| Service | July | Now | What moved |
|---|---|---|---|
| TemPolor | not in the line-up | v4.8 (songs) + i4 (instrumentals) | new to this comparison, and two new models |
| Mureka | V9 | V9.5 | new default model |
| Flow Music | Lyria 3 Pro | Lyria 3.5 | new flagship — but 3 Pro is still the API default |
| MiniMax | music-2.6 via PixVerse | Music-3.0, native | new model and a new endpoint |
Two of those are ordinary version bumps. The MiniMax row is not: MiniMax music is now a first-class endpoint rather than something you reach through PixVerse. Different endpoint, different billing.
TemPolor is the other exception. It is the only service here that splits the job across two endpoints with two different models: POST music/song runs v4.8 for anything with vocals, and POST music/instrumental runs i4 for anything without. Older model names still work and are routed to the new ones automatically, so existing code keeps running — but it also means you can be on the new models without noticing, or miss them by passing nothing at all.
The lineup and what it costs
| Model | Through useapi.net | The vendor's own API |
|---|---|---|
| TemPolor v4.8 / i4 | $30/month, unlimited tracks — no per-track rate at all | $0.04/call in prepaid credits — $100 to $5,000 a package, no bulk discount |
| MiniMax Music-3.0 | $0.010/track | $0.15/composition, +$0.01 if it writes the lyrics — 15× |
| Google Lyria 3.5 | $0.011/track | $0.08/song — 7× |
| Mureka V9.5 | $0.0135/track | $0.03/song, and the plan starts at $1,000/month — 2× |
Per-track figures are each service's highest-volume monthly plan — MiniMax Pro, Flow Music Member, Mureka Premier. Smaller plans cost more per track and still come in well under the official APIs.
TemPolor does not fit that column. Its Ultra plan sells unlimited generations for a flat monthly fee, and our submit responses came back with "unlimited": true and no remaining-track count. Its cost per track is therefore just $30 divided by however many you make — a cent at a hundred a month, a tenth of a cent at a thousand. Prices exclude VAT.
Going direct is a different shape of cost. TemPolor's own API charges $0.04 a call for the same model, paid in credits bought up front. Those credits cost $0.01 each at every package size — $100 buys 10,000, $5,000 buys 500,000 — so volume never makes a track cheaper.
What the bigger packages buy is concurrency: five requests at a time at $100, ten at $500, twenty at $5,000, and the tier does not stack if you buy the same package twice. The Ultra subscription this article ran on allows ten concurrent generations, unlimited, for $30 a month.
One naming quirk if you compare the two routes: the model this article calls v4.8 is tempolor-latest there. i4 keeps its name.
These services all sell two things: a consumer subscription, priced per month with a generous allowance, and an official API, priced per call or per credit. The API rate is several times the effective per-song rate of the subscription, because that is where the margin lives.
useapi.net connects to the account you already have. You subscribe to TemPolor, Mureka, MiniMax or Flow Music the way any user would, and we expose that account as a REST API — so generation is billed at consumer-subscription rates, not API rates. On top of that, useapi.net is a flat $15/month for every service we support, with no per-generation surcharge.
Every call in this article used a single token against all four services — no per-service SDK, no separate API keys, no four sets of billing to reconcile. The same token reaches every other service we support: video, image, speech, face swap.
How we tested
Same six briefs as July, copied from the published article rather than retyped — two instrumentals, two vocal tracks where the model writes its own lyrics, and two where we supply both the lyrics and an elaborate style brief. Every model in a round receives the identical text, so what you hear is the model. TemPolor's bodies were copied byte for byte from the MiniMax ones, which take the same prompt and lyrics fields.
Generation times are wall-clock through the API at 15-second polling granularity, so treat them as ballpark rather than benchmark. Where a model titles its own output we kept the title, spelling and stray spaces included. Tracks are complete and unedited, exactly as each API returned them: Mureka two songs per generation as 192 kbps MP3, MiniMax one at 256 kbps, TemPolor one at 48 kHz and roughly 230–320 kbps, and Flow Music an A/B pair as lossless 48 kHz WAV, transcoded to 320 kbps MP3 only so a browser can play it. Flow's lossless originals come straight from the API.
One thing to expect if you build against TemPolor: a song job reports status_final: true and COMPLETED a few seconds before its file is fetchable, and GET music/download/job_id answers 404 in the gap. Three of our four vocal rounds hit it on the first attempt. Treat that 404 as "not yet" and keep polling rather than as a failure.
Part 1 — Instrumentals
Round 1: big-band noir
Big-band noir action theme, around 60 seconds. Walking upright-bass and
baritone-sax groove, snapping rimshot snare, stabbing trumpet-section hits,
vibraphone glints, surf-twang guitar riff over a chromatic minor progression,
roaring full-brass climax with crash cymbals. Retro secret-agent swagger,
punchy analog-style production.
TemPolor i4 — "Generic Rhythm Tune" · 3:09 · generated in 103 s
curl --location 'https://api.useapi.net/v1/tempolor/music/instrumental' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model_instrumental="i4"' \
--form 'prompt="Big-band noir action theme, around 60 seconds. Walking upright-bass and baritone-sax groove, snapping rimshot snare, stabbing trumpet-section hits, vibraphone glints, surf-twang guitar riff over a chromatic minor progression, roaring full-brass climax with crash cymbals. Retro secret-agent swagger, punchy analog-style production."'
Mureka V9.5 — "Brass Knuckle Boulevard" · 3:19 · generated in 86 s
curl --location 'https://api.useapi.net/v1/mureka/music/create-instrumental' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="V9.5"' \
--form 'prompt="Big-band noir action theme, around 60 seconds. Walking upright-bass and baritone-sax groove, snapping rimshot snare, stabbing trumpet-section hits, vibraphone glints, surf-twang guitar riff over a chromatic minor progression, roaring full-brass climax with crash cymbals. Retro secret-agent swagger, punchy analog-style production."'
MiniMax Music-3.0 — "Midnight Operative" · 2:51 · generated in 182 s
curl --location 'https://api.useapi.net/v1/minimax/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="music-3.0"' \
--form 'prompt="Big-band noir action theme, around 60 seconds. Walking upright-bass and baritone-sax groove, snapping rimshot snare, stabbing trumpet-section hits, vibraphone glints, surf-twang guitar riff over a chromatic minor progression, roaring full-brass climax with crash cymbals. Retro secret-agent swagger, punchy analog-style production."' \
--form 'instrumental="true"'
Lyria 3.5 — "Double Cross" · 2:52 · generated in 47 s
curl --location 'https://api.useapi.net/v1/flowmusic/music' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="lyria-3.5"' \
--form 'instrumental="true"' \
--form 'prompt="Big-band noir action theme, around 60 seconds. Walking upright-bass and baritone-sax groove, snapping rimshot snare, stabbing trumpet-section hits, vibraphone glints, surf-twang guitar riff over a chromatic minor progression, roaring full-brass climax with crash cymbals. Retro secret-agent swagger, punchy analog-style production."'
The July versions of this brief are on the previous page if you want them back to back.
Round 2: liquid drum and bass
Liquid drum and bass instrumental, around 60 seconds. Fast rolling breakbeats
at 174 BPM, deep warm sub-bass, lush atmospheric pads, delicate piano motif,
airy wordless vocal-chop textures, smooth builds into a euphoric flowing drop.
Silky, melodic, late-night energy, pristine modern production.
TemPolor i4 — " Generic Rhythm Beat" · 3:30 · generated in 69 s
curl --location 'https://api.useapi.net/v1/tempolor/music/instrumental' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model_instrumental="i4"' \
--form 'prompt="Liquid drum and bass instrumental, around 60 seconds. Fast rolling breakbeats at 174 BPM, deep warm sub-bass, lush atmospheric pads, delicate piano motif, airy wordless vocal-chop textures, smooth builds into a euphoric flowing drop. Silky, melodic, late-night energy, pristine modern production."'
Mureka V9.5 — "Piano Afterglow" · 3:22 · generated in 85 s
curl --location 'https://api.useapi.net/v1/mureka/music/create-instrumental' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="V9.5"' \
--form 'prompt="Liquid drum and bass instrumental, around 60 seconds. Fast rolling breakbeats at 174 BPM, deep warm sub-bass, lush atmospheric pads, delicate piano motif, airy wordless vocal-chop textures, smooth builds into a euphoric flowing drop. Silky, melodic, late-night energy, pristine modern production."' \
--form 'title="Take B - Liquid DnB"'
MiniMax Music-3.0 — "Midnight Flow" · 3:45 · generated in 227 s
curl --location 'https://api.useapi.net/v1/minimax/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="music-3.0"' \
--form 'prompt="Liquid drum and bass instrumental, around 60 seconds. Fast rolling breakbeats at 174 BPM, deep warm sub-bass, lush atmospheric pads, delicate piano motif, airy wordless vocal-chop textures, smooth builds into a euphoric flowing drop. Silky, melodic, late-night energy, pristine modern production."' \
--form 'instrumental="true"'
Lyria 3.5 — "Silken Tides" · 1:06 · generated in 59 s
curl --location 'https://api.useapi.net/v1/flowmusic/music' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="lyria-3.5"' \
--form 'prompt="Liquid drum and bass instrumental, around 60 seconds. Fast rolling breakbeats at 174 BPM, deep warm sub-bass, lush atmospheric pads, delicate piano motif, airy wordless vocal-chop textures, smooth builds into a euphoric flowing drop. Silky, melodic, late-night energy, pristine modern production."' \
--form 'instrumental="true"'
None of the four APIs exposes a duration parameter — every brief asked for "around 60 seconds" in prose. Lyria was the only one that ever took the hint, and only here. Lyria 3 Pro answered this same prompt in July at 2:46.
TemPolor's instrumental endpoint also does not really name its output. It returned "Generic Rhythm Tune" and "Generic Rhythm Beat" for these two briefs — placeholders rather than titles, and not even tidy ones: the first arrives with a double space in the middle, the second with a leading space. Its song endpoint, on the next four rounds, titles properly. If you file tracks by title, generate the name yourself for instrumentals and trim whatever comes back.
Part 2 — Vocals, model writes the lyrics
Round 3: pop duet
One sentence of direction, and a deliberate capability probe — a duet only works if the model can stage two voices.
A playful pop duet between a male and a female singer about texting too late
at night, around 60 seconds, catchy back-and-forth chorus.
TemPolor v4.8 — "Text Me" · 1:51 · generated in 70 s
curl --location 'https://api.useapi.net/v1/tempolor/music/song' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model_song="v4.8"' \
--form 'prompt="A playful pop duet between a male and a female singer about texting too late at night, around 60 seconds, catchy back-and-forth chorus."'
Listen to the first line. TemPolor wrote its own lyrics for this brief, and wrote them with the speaker tagged inline:
[Verse]
[Male] You sent "you up?" at 1:09
[Female] You said "just one more," then took your time
[Male] I saw your typing, then it slipped away
[Female] I dropped my phone and missed your "hey"
Its singer strips [Verse] and [Chorus] as structure, but an inline [Male] is not a tag it recognises — so it sings the words "male" and "female" as part of the line. The two halves of the same model disagree about its own format. MiniMax met exactly the same problem on this brief and got away with it by writing (Male) and (Female) in round brackets, which its singer does drop.
The lesson generalises past this one model. Role labels inside lyrics are not a portable API feature, so if you are supplying lyrics yourself and want two voices, check what a given model does with your bracket style before you trust it — or leave the labels out and describe the duet in the style prompt instead.
Mureka V9.5 — "Are You Still Awake?" · 3:05 · generated in 88 s
curl --location 'https://api.useapi.net/v1/mureka/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="V9.5"' \
--form 'prompt="A playful pop duet between a male and a female singer about texting too late at night, around 60 seconds, catchy back-and-forth chorus."'
MiniMax Music-3.0 — "Three Little Dots" · 2:31 · generated in 146 s
curl --location 'https://api.useapi.net/v1/minimax/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="music-3.0"' \
--form 'prompt="A playful pop duet between a male and a female singer about texting too late at night, around 60 seconds, catchy back-and-forth chorus."'
MiniMax named it after the typing indicator, and its (Male) / (Female) tags stayed out of the vocal.
Lyria 3.5 — "You Go First" · 1:59 · generated in 74 s
curl --location 'https://api.useapi.net/v1/flowmusic/music' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="lyria-3.5"' \
--form 'prompt="A playful pop duet between a male and a female singer about texting too late at night, around 60 seconds, catchy back-and-forth chorus."'
Lyria wrote its own two-voice call and response, with the answering lines in parentheses.
Round 4: Spanish flamenco-pop
The language probe — the prompt is English, the song should not be:
A fiery flamenco-pop song in Spanish about summer rain, around 60 seconds,
passionate female vocals.
TemPolor v4.8 — "Lluvia" · 2:58 · generated in 102 s
curl --location 'https://api.useapi.net/v1/tempolor/music/song' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model_song="v4.8"' \
--form 'prompt="A fiery flamenco-pop song in Spanish about summer rain, around 60 seconds, passionate female vocals."'
Mureka V9.5 — "Agua de Julio" · 3:32 · generated in 87 s
curl --location 'https://api.useapi.net/v1/mureka/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="V9.5"' \
--form 'prompt="A fiery flamenco-pop song in Spanish about summer rain, around 60 seconds, passionate female vocals."'
MiniMax Music-3.0 — "Tacones de Lluvia" · 3:07 · generated in 171 s
curl --location 'https://api.useapi.net/v1/minimax/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="music-3.0"' \
--form 'prompt="A fiery flamenco-pop song in Spanish about summer rain, around 60 seconds, passionate female vocals."'
Lyria 3.5 — "Lluvia de Verano" · 2:33 · generated in 73 s
curl --location 'https://api.useapi.net/v1/flowmusic/music' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="lyria-3.5"' \
--form 'prompt="A fiery flamenco-pop song in Spanish about summer rain, around 60 seconds, passionate female vocals."'
All four wrote Spanish lyrics from an English prompt, and all four titled the track in Spanish without being asked. TemPolor's v4.8 is documented as covering 30+ languages. One Spanish brief does not test thirty of them, but it answered this one in the right language and named it "Lluvia" unprompted.
Part 3 — Vocals, our lyrics and an elaborate style brief
For the final format each model receives the same original lyrics plus a detailed production brief. Mureka takes these through music/create-advanced (lyrics + a comma-separated style desc + vocal_gender), while MiniMax, Flow Music and TemPolor all take a lyrics field next to the style prompt, so the same body works on three of the four unchanged.
Round 5: deep house
[Verse]
Two shadows in a velvet room
Moving slow to a midnight tune
Say it soft, we do not need more
The bass is talking through the floor
[Chorus]
Keep it low light, low light, hold me near
All we need is the bassline here
Round and round till the morning shows
Low light, low light, do not let go
(do not let go, do not let go)
Deep house at 121 BPM. Rolling sub-bass groove, shuffled hi-hats, warm Rhodes
chord stabs, subtle chopped vocal echoes, smoky intimate female lead vocal,
understated hypnotic late-night club production.
TemPolor v4.8 — "Low Light" · 1:28 · generated in 51 s
curl --location 'https://api.useapi.net/v1/tempolor/music/song' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model_song="v4.8"' \
--form 'prompt="Deep house at 121 BPM. Rolling sub-bass groove, shuffled hi-hats, warm Rhodes chord stabs, subtle chopped vocal echoes, smoky intimate female lead vocal, understated hypnotic late-night club production."' \
--form 'lyrics="[Verse]
Two shadows in a velvet room
Moving slow to a midnight tune
Say it soft, we do not need more
The bass is talking through the floor
[Chorus]
Keep it low light, low light, hold me near
All we need is the bassline here
Round and round till the morning shows
Low light, low light, do not let go
(do not let go, do not let go)"'
Mureka V9.5 — "Low Light" · 3:09 · generated in 101 s
curl --location 'https://api.useapi.net/v1/mureka/music/create-advanced' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="V9.5"' \
--form 'lyrics="[Verse]
Two shadows in a velvet room
Moving slow to a midnight tune
Say it soft, we do not need more
The bass is talking through the floor
[Chorus]
Keep it low light, low light, hold me near
All we need is the bassline here
Round and round till the morning shows
Low light, low light, do not let go
(do not let go, do not let go)"' \
--form 'desc="deep house, late night, smoky, hypnotic, groovy, 121 bpm, female vocals, warm, understated club production"' \
--form 'vocal_gender="female"'
MiniMax Music-3.0 — "Low Light" · 2:47 · generated in 135 s
curl --location 'https://api.useapi.net/v1/minimax/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="music-3.0"' \
--form 'prompt="Deep house at 121 BPM. Rolling sub-bass groove, shuffled hi-hats, warm Rhodes chord stabs, subtle chopped vocal echoes, smoky intimate female lead vocal, understated hypnotic late-night club production."' \
--form 'lyrics="[Verse]
Two shadows in a velvet room
Moving slow to a midnight tune
Say it soft, we do not need more
The bass is talking through the floor
[Chorus]
Keep it low light, low light, hold me near
All we need is the bassline here
Round and round till the morning shows
Low light, low light, do not let go
(do not let go, do not let go)"'
Lyria 3.5 — "Velvet Room" · 2:26 · generated in 43 s
curl --location 'https://api.useapi.net/v1/flowmusic/music' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="lyria-3.5"' \
--form 'prompt="Deep house at 121 BPM. Rolling sub-bass groove, shuffled hi-hats, warm Rhodes chord stabs, subtle chopped vocal echoes, smoky intimate female lead vocal, understated hypnotic late-night club production."' \
--form 'lyrics="[Verse]
Two shadows in a velvet room
Moving slow to a midnight tune
Say it soft, we do not need more
The bass is talking through the floor
[Chorus]
Keep it low light, low light, hold me near
All we need is the bassline here
Round and round till the morning shows
Low light, low light, do not let go
(do not let go, do not let go)"'
Round 6: jazzy house duet
The hardest brief of the set — a two-voice arrangement with live-jazz layers over a house groove, voice roles tagged in the lyrics. The tags are a hint, not an API parameter, so listen for who actually trades verses and who flattens it into one singer.
[Verse 1 - male]
Blue hour settles on the avenue
Trumpet smoke and a table for two
I talk in rhythm, you answer in rhyme
We swing together on borrowed time
[Verse 2 - female]
You bring the thunder of a double bass
I bring the shimmer of a satin dress
Piano is asking what the night is about
We are the answer when the lights go out
[Chorus - both]
Meet me in the blue hour, where the bass goes slow (blue hour, blue hour)
Sax on the corner and the lights burn low (burn low, burn low)
You take the high note, I will hold the ground
Blue hour, blue hour - do not stop the sound
TemPolor v4.8 — "Blue Hour" · 2:09 · generated in 69 s
curl --location 'https://api.useapi.net/v1/tempolor/music/song' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model_song="v4.8"' \
--form 'prompt="Jazzy deep house duet at 120 BPM. Rolling deep-house groove with swung hi-hats, walking upright-bass motif over the sub, smoky muted trumpet and tenor-sax licks answering the vocals, extended jazz piano chords, a male and a female lead trading verses and joining in harmony on the chorus, small backing ensemble echoing the hook, warm vinyl texture, sophisticated late-night club production."' \
--form 'lyrics="[Verse 1 - male]
Blue hour settles on the avenue
Trumpet smoke and a table for two
I talk in rhythm, you answer in rhyme
We swing together on borrowed time
[Verse 2 - female]
You bring the thunder of a double bass
I bring the shimmer of a satin dress
Piano is asking what the night is about
We are the answer when the lights go out
[Chorus - both]
Meet me in the blue hour, where the bass goes slow (blue hour, blue hour)
Sax on the corner and the lights burn low (burn low, burn low)
You take the high note, I will hold the ground
Blue hour, blue hour - do not stop the sound"'
Mureka V9.5 — "Blue Hour" · 3:12 · generated in 70 s
curl --location 'https://api.useapi.net/v1/mureka/music/create-advanced' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="V9.5"' \
--form 'lyrics="[Verse 1 - male]
Blue hour settles on the avenue
Trumpet smoke and a table for two
I talk in rhythm, you answer in rhyme
We swing together on borrowed time
[Verse 2 - female]
You bring the thunder of a double bass
I bring the shimmer of a satin dress
Piano is asking what the night is about
We are the answer when the lights go out
[Chorus - both]
Meet me in the blue hour, where the bass goes slow (blue hour, blue hour)
Sax on the corner and the lights burn low (burn low, burn low)
You take the high note, I will hold the ground
Blue hour, blue hour - do not stop the sound"' \
--form 'desc="jazzy deep house, jazz house, duet, male and female vocals, group backing vocals, muted trumpet, saxophone, upright bass, swung, 120 bpm, late night, warm sophisticated production"'
MiniMax Music-3.0 — "Blue Hour" · 2:18 · generated in 134 s
curl --location 'https://api.useapi.net/v1/minimax/music/create' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="music-3.0"' \
--form 'prompt="Jazzy deep house duet at 120 BPM. Rolling deep-house groove with swung hi-hats, walking upright-bass motif over the sub, smoky muted trumpet and tenor-sax licks answering the vocals, extended jazz piano chords, a male and a female lead trading verses and joining in harmony on the chorus, small backing ensemble echoing the hook, warm vinyl texture, sophisticated late-night club production."' \
--form 'lyrics="[Verse 1 - male]
Blue hour settles on the avenue
Trumpet smoke and a table for two
I talk in rhythm, you answer in rhyme
We swing together on borrowed time
[Verse 2 - female]
You bring the thunder of a double bass
I bring the shimmer of a satin dress
Piano is asking what the night is about
We are the answer when the lights go out
[Chorus - both]
Meet me in the blue hour, where the bass goes slow (blue hour, blue hour)
Sax on the corner and the lights burn low (burn low, burn low)
You take the high note, I will hold the ground
Blue hour, blue hour - do not stop the sound"'
Lyria 3.5 — "Blue Hour" · 2:34 · generated in 58 s
curl --location 'https://api.useapi.net/v1/flowmusic/music' \
--header 'Authorization: Bearer user:12345-…' \
--form 'model="lyria-3.5"' \
--form 'prompt="Jazzy deep house duet at 120 BPM. Rolling deep-house groove with swung hi-hats, walking upright-bass motif over the sub, smoky muted trumpet and tenor-sax licks answering the vocals, extended jazz piano chords, a male and a female lead trading verses and joining in harmony on the chorus, small backing ensemble echoing the hook, warm vinyl texture, sophisticated late-night club production."' \
--form 'lyrics="[Verse 1 - male]
Blue hour settles on the avenue
Trumpet smoke and a table for two
I talk in rhythm, you answer in rhyme
We swing together on borrowed time
[Verse 2 - female]
You bring the thunder of a double bass
I bring the shimmer of a satin dress
Piano is asking what the night is about
We are the answer when the lights go out
[Chorus - both]
Meet me in the blue hour, where the bass goes slow (blue hour, blue hour)
Sax on the corner and the lights burn low (burn low, burn low)
You take the high note, I will hold the ground
Blue hour, blue hour - do not stop the sound"'
With the lyrics supplied, all four took the title from the chorus hook. Note that this round's role labels sit inside the section headers — [Verse 1 - male] — rather than on their own at the head of a line, which is the form models treat as structure.
FAQ
What does a track cost through useapi.net? On three of the four, about a cent: MiniMax $0.010 a track, Lyria through Flow Music $0.011, and Mureka $0.0135, each on that service's highest-volume monthly plan. TemPolor has no per-track rate — its Ultra plan is a flat $30 a month for unlimited generations, so the more you generate the less each track costs.
Which one is cheapest depends on volume, then? Yes. Like for like — the same TemPolor model on its own API at $0.04 a call — the flat $30 plan breaks even at 750 tracks a month and wins above it. Because TemPolor's credits cost the same at every package size, that break-even does not move as you scale up. Against the other services here, $30 buys roughly 2,200 tracks at Mureka's rate or 3,000 at MiniMax's. Add useapi.net's flat $15/month once, whichever you pick, and nothing per generation.
Why is it cheaper than the official APIs? Because you bring your own account. useapi.net exposes the consumer subscription you already pay for as a REST API, so generation is billed at subscription rates rather than per-call API rates. useapi.net itself is a flat $15/month for every supported service, with no per-generation surcharge.
Can I run all four from one account? Yes. One useapi.net token reaches TemPolor, Mureka, MiniMax and Flow Music — every call in this article was made with a single token — plus every other service we support. You still hold your own account with each service.
How do I pick TemPolor's newest models? They are already the defaults: v4.8 on POST music/song via model_song, and i4 on POST music/instrumental via model_instrumental. Older model names are still accepted and routed to the new ones, so you cannot accidentally stay on an old model — but pass the parameter explicitly if you want your code to state which model produced a track.
Why does TemPolor have two endpoints when the others have one? Because vocals and instrumentals are different models there — v4.8 and i4. Pick the endpoint by whether you want singing, not by a flag. The other three take an instrumental parameter on a single endpoint instead.
How do I pick Lyria 3.5 rather than 3 Pro? Pass model: "lyria-3.5" on POST /music. lyria-3-pro remains the default, so omitting the parameter renders the older model.
Do any of these APIs let me set the track length? No. None of the four exposes a duration parameter. The amount of lyrics you supply is the only lever — more words generally means a longer song.
Can I get more than one track per call? On MiniMax, yes — quantity up to 3, billed as three songs. Mureka returns two songs per generation and Flow Music an A/B pair, both included in the one generation. TemPolor returned one track per call throughout this test, and allows up to ten generations running at once.
Can I put role labels like [Male] in the lyrics I send? You can send them, but no API here treats them as a parameter, and what happens next differs per model — TemPolor sings inline [Male] out loud, MiniMax drops (Male). Describe the duet in the style prompt and test your bracket style before relying on it. See round 3.
Top comments (0)