Custom e-learning development in 2026 costs roughly $83 to $500 per finished minute of course video, based on the handful of vendors who publish rates. A standard 15-minute module therefore quotes anywhere from about $1,250 to $7,500, and project minimums at established shops commonly start at $10,000.
That spread is not noise. It maps to specific, checkable variables — and because most vendors publish nothing, the few who do are the only honest reference points a buyer has. Here they are.
The published numbers
| Source | Published rate | Normalized |
|---|---|---|
| Industry consensus floor (widely cited) | $5,000 per finished hour | ≈ $83/min |
| TTMS — published price list, 2025 | €1,622 per 15-min standard module | ≈ $117/min |
| Blue Carrot — published guide, 2026 | $2,500–3,500 per 15-min compliance module | ≈ $167–233/min |
| LearningSim — published, 2024 | Level-1 courses $12,000–15,000 per finished hour | ≈ $200–250/min |
| Ninja Tropic — published own rates | Host-led video $250–500 per minute | $250–500/min |
| Clutch-listed e-learning shops | Project minimums | $10,000–50,000 |
Two cautions before you anchor on any of these. First, “finished minute” is the only unit that compares fairly — beware quotes per slide, per page of source material, or per “learning hour”, which can hide 2–3× differences. Second, regional markets diverge sharply: Indonesian local vendors, for example, have quoted around Rp 30–50 million per learning hour (roughly $310–520 per 10 minutes), a figure we class as anecdote-grade because no Indonesian production house we could find publishes a rate card.
The four drivers behind the 6× spread
Finished runtime. The strongest driver in every published rate structure. Costs scale with output minutes, not input pages — and good instructional scripting usually compresses classroom seat time substantially, which means the cheapest minute is the one a better script removed.
Interactivity and assessment. This is what separates the $83 floor from $250+ boutique rates, far more than video polish does. Passive video sits at the bottom of the range; embedded knowledge checks, scenario branching, and completion gating climb it. LearningSim’s published tiers make this explicit: their “Level 1” is the simple end, and each interactivity level above it steps the price.
Languages. A second language version does not cost a second course — structure, design, and assessments carry over. Itemized per-language pricing (a fraction of the master course per language) is the pattern to ask for; a vendor who quotes each language at full price is reselling you work they already did.
Review cycles. Script-stage changes cost minutes; post-production changes cost re-rendering or, at traditional shops, re-shooting. Vendors price this risk in. A statement of work that names its included revision rounds — and a client who concentrates review at the script stage — lands at the bottom of any vendor’s range.
What a fair quote looks like
A credible quote in this market is a fixed, itemized statement of work: modules, finished runtime, interactivity level, languages, revision rounds, timeline, one price. Hourly estimates and “ranges subject to discovery” shift scoping risk onto you. The published-rate vendors above all quote fixed; insist on the same from anyone who doesn’t publish.
One business day is long enough to produce such a quote from representative material. Longer usually means the vendor is hand-building an estimate they will defend later, instead of running a scoping process they trust.
The full benchmark — every source above with citations, a vendor-evaluation checklist, and the anatomy of a good SOW — is available as a downloadable buyer’s benchmark. For how Voctus itself prices work, see Pricing & scoping.