Quso.ai credits at about one per minute vs Sume priced by step
Quso.ai says one credit is roughly one minute processed, from 100 credits for $29. Sume prices steps: $0.01 transcript, $0.02 trim, $0.10 render, $0.20 caption.

On quso.ai a credit is about one minute of processed video, and on the monthly-billed Lite plan that is $0.29 a credit, falling to about $0.03 on the annual Growth plan. Sume has no credits: a transcript is $0.01 a minute, a trim is $0.02, a render is $0.10 a minute and a 60-second caption job is $0.20. Which is cheaper depends on how many steps one minute of your video passes through.
quso.ai's pricing page (read 2026-10-03) states: "One credit ≈ one minute of processed video. Clipping, captioning, and editing draw from your monthly credits." With TikTok reportedly rewarding videos of 3 to 5 minutes, per Metricool (read 2026-10-03), one finished episode can mean several processed minutes.
What the page lists
Four plans on monthly billing: Free at $0 with 75 credits, 720p and TikTok publishing with 7-day retention; Lite at $29 with 100 credits; Essential at $39 with 300 credits and scheduling to 7 platforms; Growth at $49 with 600 credits, custom branding and analytics. Annual billing is about 33 percent cheaper per month and doubles the credits: Lite $19, Essential $26, Growth $33. The text read listed no clip limits and no API price.
| Plan | Monthly fee | Monthly credits | Per credit, monthly | Annual fee per month | Annual credits | Per credit, annual |
|---|---|---|---|---|---|---|
| Lite | $29 | 100 | $0.290 | $19 | 200 | $0.095 |
| Essential | $39 | 300 | $0.130 | $26 | 600 | $0.043 |
| Growth | $49 | 600 | $0.082 | $33 | 1,200 | $0.028 |
How Sume prices the same minute
A single "processed minute" on a credit system hides several operations. On Sume they are separate lines. Video inspect transcribes at $0.01 per audio minute on top of its Modal compute. Video trim costs $0.02 a job regardless of length. Timeline 1.0 is $0.10 per output minute. Video captions is $0.20 per video up to 60 seconds.
So transcribing a minute of source costs $0.01 and each clip you cut adds $0.02. A one-minute clip that you transcribe, trim, render and caption comes to about $0.33.
PLANS = {"Lite": (29, 100, 19, 200), "Essential": (39, 300, 26, 600), "Growth": (49, 600, 33, 1200)}
for name, (m_fee, m_cr, a_fee, a_cr) in PLANS.items():
print(f"{name}: monthly ${m_fee / m_cr:.3f} per credit, annual ${a_fee / a_cr:.3f} per credit")
print("Sume per processed minute, by step:")
for step, price in (("transcript", 0.01), ("trim", 0.02), ("render", 0.10), ("caption (60 s)", 0.20)):
print(f" {step}: ${price:.2f}")
Worked examples
A podcast producer uploads a 60-minute episode and wants six 90-second clips with captions. On quso.ai that is at least 60 credits to process the source, plus whatever the clips draw; on the annual Essential plan that is under $3 of credits at $0.043 each. On Sume the transcript is $0.60 across two inspect jobs, each 90-second clip is cut as two 45-second parts so that each part fits the 60-second caption limit, which is twelve trims at $0.24 and twelve caption jobs at $2.40, and six renders to join the parts at $0.20 each (two minutes rounded up) are $1.20. The total is $4.44.
That example shows where Sume is worse: a source-minute credit plan is very cheap when the source is long and the output is short. Reverse the shape. A creator makes 3-minute episodes from the same 3-minute footage and does only a render and a caption pass. If each step draws one credit per minute, as the page's wording suggests, that is 3 to 9 credits. Sume is 3 render minutes at $0.10 plus 3 caption parts at $0.20, $0.90. A Lite plan on monthly billing at $0.29 per credit is $0.87 to $2.61 for the same.
Sume's per-step lines are easy to predict because each step has a listed price, and the Timeline plan call returns the render cost before you spend.
What Sume does and does not do here
- Sume does not clip automatically or schedule posts. quso.ai lists scheduling to 7 platforms on Essential.
- Sume gives you a price for each step before it runs, either from the live catalog or the unbilled Timeline plan call.
- Sume steps are separate jobs, each with an
Idempotency-Key, so a failed caption job does not force a repeat render. - Sume's job output is an MP4 URL; there is no hosted workspace to log into.
What to do
Pick the step you use most. If you mostly clip long sources and rarely caption, an annual credit plan is hard to beat. If you mostly render and caption known edits, price those two lines on Sume and compare.
Sources
Related posts
More in Comparisons
- Qwen Image vs Qwen Image Max on the Sume image API
Qwen Image lists $0.025 and edits with up to 10 references; Qwen Image Max lists $0.09375 and is text-to-image only. Catalog table and request examples.
- Record a voice sample in the browser: Sume Voices 30 seconds
Sume Voices clone mode takes an uploaded clip or a browser recording of up to 30 seconds. ElevenLabs states clone samples from 10 seconds to two minutes.
- Replicate MCP discovery via server.json vs Sume's MCP URL
Replicate publishes /.well-known/mcp/server.json for the official MCP Registry. Sume documents one hosted MCP URL and OAuth metadata. How each client connects.
- Replicate allow_fallback_model: Nano Banana Pro vs Sume
Replicate can fall back from Nano Banana Pro to Seedream 5.0 lite and bills the fallback. Sume's allow_fallbacks is accepted but has no effect. What to do.
Written by Sume