Understanding TubeOnAI’s Credit Billing System
Introduction
TubeOnAI uses a credit-based billing system so you get transparent pricing based on what you use. Whether you summarize
a 3-minute YouTube clip or a 2-hour podcast, credits show what each operation costs. You can also choose the AI model
that fits your budget.
This guide covers how credits work, what they cost for each content type, how to pick the right AI model, and a detailed
breakdown of every subscription tier, including AppSumo lifetime deals.
All numbers in this article are based on real production data from 56,700+ completed operations across 12,250+ users
over the last 90 days.
Why credits?
Before credits, TubeOnAI used fixed monthly limits, like “100 summaries per month,” even if one summary was a 2-minute
clip and another was a 2-hour lecture. That was not fair.
Credits fix this by matching the real AI cost of each operation:
- Short content costs less: a 3-minute video uses about 13 credits, while a 60-minute video uses about 51 credits.
- Cheaper models cost less: gpt-5-nano uses about 5 credits for a 60-minute video, while Claude Sonnet 4.5 uses about
96 credits.
- You choose: pick the model and content that match your needs and budget.
How credits work
Every AI operation on TubeOnAI follows a simple formula:
Credits = Transcription cost + AI summarization cost
- Transcription cost depends on whether free captions are available (YouTube, Facebook) or paid transcription is
needed (uploaded audio/video, podcasts).
- AI summarization cost depends on the model you choose and the length of the content.
1 credit = $0.001 of AI processing cost. So 100 credits = $0.10, and 1,000 credits = $1.00. Credits are deducted from
your monthly balance when each operation completes. Failed operations are automatically refunded.
Credit usage by content type
Based on 56,700+ real operations (all models averaged, default model: gpt-5-mini):
YouTube videos
| Video duration | Avg credits | With gpt-5-nano (approx) | | ---- | | Under 3 min | 13 | ~3 | | 3–10 min | 21 | ~5 |
| 10–20 min | 33 | ~6 | | 20–30 min | 40 | ~7 | | 30–60 min | 51 | ~8 | | 60–120 min | 69 | ~10 |
Key insight: YouTube videos are the cheapest content type because they use free API captions instead of paid
transcription.
Documents
| Document type | Avg credits | | ---- | | Web Article | 15 | | PowerPoint | 23 | | Document (Word/Text) | 29 | | PDF
| 38 |
Audio & Podcasts
Audio and podcast content requires paid transcription, which adds an average of 177 credits to the base summarization
cost. A 60-minute podcast typically costs 180–250 credits, depending on the AI model you choose.
Free Captions vs. Paid Transcription
This is the single biggest factor in credit cost that most users do not expect.
Free captions (low cost)
- YouTube videos: captions pulled from YouTube’s API at zero cost
- Facebook videos: captions pulled from Facebook’s API These sources only pay for AI summarization, so they are very
cheap.
Paid transcription (higher cost)
- Uploaded audio/video files (MP3, WAV, MP4)
- Podcasts (Spotify, Apple Podcasts, RSS feeds)
- Any content without existing captions Paid transcription adds an average of 177 credits per operation, on top of the
summarization cost. Example: A 60-minute YouTube video costs ~5 credits with gpt-5-nano. The same 60-minute content
uploaded as an MP3 costs ~182 credits because transcription is required.
Cost by AI Model
Different AI models have very different costs. Here is what a 60-minute YouTube video costs by model (real production
data): | AI model | Credits for a 60-min YouTube | Relative cost | | ---- | | gpt-5-nano | 5 | Baseline | | gpt-5-mini
| 16 | 3.2× | | Gemini 2.5 Flash | 19–29 | 4–6× | | gpt-5 | 68 | 14× | | Claude Sonnet 4.5 | 96 | 19× | | gpt-5.1 | 107
| 21× |
Which model should you choose?
gpt-5-nano — Best for high-volume use. Fast, low cost, and solid quality for straightforward content. Ideal for news,
short videos, and simple articles. gpt-5-mini — Best balance of quality and cost. This is the default model and works
well for most content types. Gemini 2.5 Flash — A strong alternative to gpt-5-mini with competitive pricing. gpt-5 /
Claude Sonnet 4.5 — Premium quality for complex, technical, or long-form content where nuance matters. gpt-5.1 — Highest
quality available. Use for critical content where accuracy is paramount.
Repurpose & Custom Prompts
Repurpose transforms an existing summary into a new format: blog post, Twitter thread, LinkedIn post, email newsletter,
bullet points, and more. Since the content is already processed, repurpose only charges for the AI generation step.
Repurpose cost by model (real production data)
| AI model | Avg credits per repurpose | | ---- | | gpt-5-nano | 4 | | gpt-5-mini | 18 | | gpt-5.1 | 38 | | Claude
Sonnet 4.5 | 103 |
Custom Prompts
Custom prompts work similarly to Repurpose. You provide your own instruction (for example, “Extract all action items” or
“Summarize for a 5th grader”), and the AI processes your existing content. Costs are comparable to Repurpose for the
same model. Tip: If you repurpose the same content with the same prompt and model again, TubeOnAI serves it from cache
at minimal cost.
Using a 10-minute YouTube video as the benchmark:
| Plan | Credits | gpt-5-nano (~5 cr/video) | gpt-5-mini (~21 cr/video) | gpt-5 (~68 cr/video) | | ---- | | Free | 50 |
~10 videos | ~2 videos | — | | Starter | 1,000 | ~200 videos | ~47 videos | ~14 videos | | Pro | 5,000 | ~1,000 videos |
~238 videos | ~73 videos | | Premium | 15,000 | ~3,000 videos | ~714 videos | ~220 videos | | Enterprise | 50,000 |
~10,000 videos | ~2,380 videos | ~735 videos | Using a 60-minute YouTube video as the benchmark: | Plan | Credits |
gpt-5-nano (~5 cr/video) | gpt-5-mini (~16 cr/video) | gpt-5 (~68 cr/video) | | ---- | | Free | 50 | ~10 videos | ~3
videos | — | | Starter | 1,000 | ~200 videos | ~62 videos | ~14 videos | | Pro | 5,000 | ~1,000 videos | ~312 videos |
~73 videos | | Premium | 15,000 | ~3,000 videos | ~937 videos | ~220 videos | | Enterprise | 50,000 | ~10,000 videos |
~3,125 videos | ~735 videos |
AppSumo Lifetime Deals FAQ
If you purchased TubeOnAI through AppSumo, this section answers the most common questions.
AppSumo Tiers at a Glance
| Feature | Tier 1 | Tier 2 | Tier 3 | | ---- | | Price (lifetime) | 59 | 149 | 379 | | Credits/month | 3,000 | 10,000
| 60,000 | | Channels | 3 | 10 | 500 | | Custom Prompts | 500 | 500 | 500 | | Dollar value/mo (AI) | 3.00 | 10.00
| 60.00 |
What’s the correlation between credits and dollar value?
Each credit equals $0.001 of AI processing cost.
- 3,000 credits (Tier 1) = $3.00/month of AI processing
- 10,000 credits (Tier 2) = $10.00/month of AI processing
- 60,000 credits (Tier 3) = $60.00/month of AI processing This is the raw AI cost TubeOnAI pays to providers like
OpenAI, Anthropic, and Google on your behalf each month.
How many videos can I summarize with 3,000 credits (Tier 1)?
It depends on the AI model you choose.
10-minute YouTube videos with 3,000 credits
- gpt-5-nano (~5 credits/video): ~600 videos
- gpt-5-mini (~16 credits/video): ~187 videos
- gpt-5 (~68 credits/video): ~44 videos | Model | Credits per video (approx) | Videos per month (approx) | | ---- | |
gpt-5-nano | ~5 | ~600 | | gpt-5-mini (default) | ~21 | ~142 | | gpt-5 | ~68 | ~44 | 60-minute YouTube videos
with 3,000 credits: | Model | Credits per video (approx) | Videos per month (approx) | | ---- | | gpt-5-nano | ~5 |
~600 | | gpt-5-mini (default) | ~16 | ~187 | | gpt-5 | ~68 | ~44 | Why is nano almost the same for 10-min
and 60-min? Because gpt-5-nano is so efficient that the cost difference between short and long content is small. In
most cases, it only changes by a few credits between short and long content.
Q: How many 60-minute videos can I process per AppSumo tier?
| Model | Tier 1 (3,000 cr) | Tier 2 (10,000 cr) | Tier 3 (60,000 cr) | | ---- | | gpt-5-nano | ~600 | ~2,000 | ~12,000
| | gpt-5-mini | ~187 | ~625 | ~3,750 | | Gemini 2.5 Flash | ~103–157 | ~344–526 | ~2,068–3,157 | | gpt-5 | ~44 | ~147 |
~882 | | Claude Sonnet 4.5 | ~31 | ~104 | ~625 | | gpt-5.1 | ~28 | ~93 | ~560 |
What about channel auto-summarization limits?
Tier 1: Up to 3 YouTube channels for automatic new-video summarization Tier 2: Up to 10 channels Tier 3: Up to 500
channels Each channel auto-summarizes new uploads using your chosen default model. Credit usage is deducted from your
monthly credits.
Do AppSumo credits roll over?
No. Credits reset each month on your billing cycle date. Unused credits do not carry over.
Can I upgrade my AppSumo tier?
Yes. You can stack codes on AppSumo to upgrade from Tier 1 → Tier 2 → Tier 3.
General FAQ
How many credits does a YouTube video use by length?
| Duration | gpt-5-nano (approx credits) | gpt-5-mini (default, approx credits) | gpt-5 (approx credits) | | ---- | | 3
min | ~3 | ~13 | ~50 | | 10 min | ~5 | ~21 | ~68 | | 20 min | ~6 | ~33 | ~80 | | 60 min | ~5 | ~51 | ~68 | | 120 min |
~10 | ~69 | ~100 | How many credits does a PDF use? | PDF size | Avg credits (approx) | | ---- | | 1–5 pages | ~15–20 |
| 10–20 pages | ~30–40 | | 50+ pages | ~50–80 |
How many credits does a web article use?
Web articles average 15 credits across all models. Short articles (under 1,000 words) typically cost 8–12 credits.
Long-form articles (5,000+ words) may cost 20–30 credits.
How many credits does repurposing to a blog post cost?
- gpt-5-nano: 4 credits
- gpt-5-mini: 18 credits See the full model breakdown in Section 7.
Do credits expire?
Yes. Credits reset monthly on your billing cycle date. Unused credits do not roll over to the next month.
Are there hidden fees?
No. Credits are the only cost beyond your subscription. There are no per-operation fees, API surcharges, or bandwidth
charges.
What happens if I run out of credits?
Your account pauses AI operations until credits reset next month. Existing summaries remain accessible. You can upgrade
your plan at any time for an immediate credit increase.
Can I use my own API key (BYOK)?
Yes. Bring Your Own Key users can connect their own OpenAI, Anthropic, or Google API keys. BYOK operations still consume
credits as a soft usage limit, but at reduced rates since TubeOnAI does not pay the AI provider directly.
What Affects Credit Usage?
Four factors determine how many credits an operation costs:
1. Content length — Longer content produces more tokens for the AI to process
2. AI model choice — Premium models cost 5–20× more than nano
3. Transcription method — Free captions (YouTube/Facebook) vs. paid transcription (audio uploads, podcasts)
4. Operation type — Summaries cost more than repurpose/custom prompts because they process raw content
5.
Smart Tips to Save Credits
1. Use gpt-5-nano for bulk processing. At ~5 credits per 60-min YouTube video, you can process thousands of videos per
month on any paid plan.
2. Use gpt-5-mini as your default. It offers the best quality-to-cost ratio for everyday use.
3. Reserve premium models for important content. Use gpt-5, Claude Sonnet 4.5, or gpt-5.1 only when quality matters
most.
4. Prefer YouTube/Facebook links over file uploads. Free captions save ~177 credits per operation compared to paid
transcription.
5. Repurpose from existing summaries. Repurpose costs 4–18 credits (nano/mini) vs. generating a new summary from
scratch.
6. Use cached results. Requesting the same repurpose format twice serves from cache at near-zero cost.
7. Summarize shorter clips when possible. A 10-min video costs ~21 credits vs. ~69 for a 120-min video (mini model).
8. Checking Your Credit Balance
You can check your credit balance at any time:
- Web app: Click your profile icon → Credits/Billing section
- Mobile app: Settings → Subscription & Credits
- API: GET /api/v1/credits/balance returns your current balance and usage
Your dashboard shows:
- Current credit balance
- Credits used this month
- Reset date
- Usage breakdown by content type and model
Automatic Credit Protection
TubeOnAI protects your credits with automatic safeguards:
- Failed operations are refunded. If transcription fails, the AI returns an error, or processing times out, reserved
credits are returned to your balance automatically.
- Duplicate detection. If you accidentally submit the same content twice within 30 seconds, the system prevents
double-charging.
- Pre-flight estimation. Before processing, the system estimates credit cost and checks your balance. If you don't
have enough credits, the operation is blocked before any charges.
- Overdraft protection. In rare cases where actual cost exceeds the estimate, the operation completes but your wallet
locks until credits reset — you're never charged extra beyond your plan.
Based on production data, the automatic refund system processes credit returns for failed operations within seconds,
ensuring you only pay for successfully completed work.
TubeOnAI vs. Doing It Yourself
How does TubeOnAI's credit pricing compare to using AI APIs directly?
| What you’d need | DIY cost | TubeOnAI | | ---- | | OpenAI API key + transcription | Pay per token + Whisper costs |
Included in credits | | Build transcription pipeline | Engineering time | Built-in | | Handle failures & retries |
Engineering time | Automatic | | Multi-model fallback | Multiple API keys + logic | Built-in | | YouTube caption
extraction | YouTube API quota + code | Built-in | | Document parsing (PDF, PPTX) | Multiple libraries + hosting |
Built-in | | Content repurposing | Custom prompts + API calls | One click |
Platform stats (last 90 days):
- 56,700+ operations completed successfully
- 12,250+ unique users served
- 2.17M total credits consumed
- 38.2 average credits per operation
At 38.2 credits ($0.038) per average operation, TubeOnAI handles transcription, summarization, formatting, storage, and
delivery — work that would require significant engineering effort and higher raw API costs to replicate independently.
Last updated: February 2026. All credit figures based on real production data from the last 90 days.