Amazon DSP Streaming Inventory: The Full List of What You Can Buy

Amazon-owned inventory

Prime Video is the anchor: full-screen non-skippable video, pause ads, and — since 2026 — interactive and dynamic creative formats, plus Amazon's own exclusive Thursday Night Football rights through 2033. Fire TV Channels adds more than 900 ad-supported channels across 30+ categories, plus device-level inventory — Feature Rotator, sponsored tiles, banners and screensavers — that sits outside any specific piece of content. Twitch contributes both display (since December 2022) and, since being folded into the unified Streaming TV product, full-screen video. What used to be Freevee's separate ad-supported catalog now lives inside Prime Video's free, ad-supported section, worth its own reporting line even though it's no longer a separate buy.

These are the properties with the longest track record inside Amazon DSP, and generally the most predictable delivery — worth treating as the base of a streaming plan that third-party inventory then extends, rather than as one option among many equally-weighted choices.

Disney's bundle, via DRAX

Amazon Ads and Disney Advertising connected Amazon DSP to Disney's Real-Time Ad Exchange on 17 June 2025 — Amazon became only the third DSP, after Google and The Trade Desk, with that access. The bundle covers Disney+, ESPN and Hulu, with Disney's Magic Words contextual targeting and, eventually, Disney Select first-party data layered on top, all matched against Amazon's own signals through the Amazon Publisher Cloud clean room. Rollout reached all US Amazon DSP advertisers in Q3 2025, after Disney+ launched first in a set of international markets.

Netflix, and the extended Roku footprint

Netflix became directly buyable through Amazon DSP starting Q4 2025, announced 10 September 2025, across roughly a dozen global markets including the US, UK, Canada, Australia, Japan, Brazil, Mexico and much of Western Europe. Separately, a June 2025 partnership with Roku extended Amazon's addressability — its ability to recognize a logged-in household and apply frequency capping and targeting consistently — across an estimated 80 million US CTV households on Roku-connected apps, more than 80% of US CTV-watching households by one industry estimate. That partnership is specifically about identity and addressability across Roku's device footprint, not a separate direct-buy integration the way Disney and Netflix are — worth keeping the distinction straight when you're planning a buy rather than reading about reach.

How big this actually is: a worked example

The number that captures the scale better than any single platform list is this: across 27 advertisers in our own Amazon DSP book over a 31-day window this summer, delivery touched 134 distinct named supply sources — not two or three placements, a genuinely fragmented landscape where the performance spread between supply sources was larger than the spread between advertisers. A brand buying "Amazon DSP streaming" without a supply-source breakdown is buying an average of things that behave nothing alike — Alexa device inventory in that same period ran a $0.32 CPM and 6.1% new-to-brand, while premium streaming inventory ran $15.51 CPM and 56.3% new-to-brand. Averaging those two into one number tells you almost nothing useful.

Put a number on the practical consequence: if $50,000 of a monthly budget is spread evenly, undifferentiated, across a mix of cheap owned-and-operated inventory and premium third-party streaming, the blended CPM and blended new-to-brand rate you report to a stakeholder describes neither piece of the buy accurately. The 134-source figure isn't a piece of trivia — it's the actual argument for reading a supply-source report before reading a blended one, on every account, every month, not just when something looks off.

What to do when the list changes on you

This list moves. Freevee disappeared as a named product between 2024 and 2025; Netflix and the Disney bundle both arrived within a few months of each other in 2025; the Roku partnership changed what "addressable" means for CTV households outside Amazon's own ecosystem entirely. If a plan built six months ago references a supply source that's since changed name or absorbed into something else, that's not a hypothetical — it's happened at least twice in this category already. Check current supply-source names in your DSP console at the start of every planning cycle rather than trusting last quarter's line-item list.

The safest habit is to pull a fresh supply-source delivery report every month rather than relying on a saved list of named platforms in a media plan document. A report shows you what actually delivered, under whatever name it currently carries — a document written last quarter shows you what someone believed was buyable at the time it was written.

The common mistake

The mistake is planning against a mental list of "the big three or four" streaming platforms and missing that the actual buyable inventory is an order of magnitude more fragmented than that, with genuinely different performance by supply source. reMKTR builds every streaming plan starting from a supply-source breakdown, not a platform list, specifically because the platform list is the thing that goes stale first — and because the 134-source spread in our own book is the strongest evidence we have that averaging across them, rather than reading each one, is where most of the wasted budget in a streaming plan actually hides.

A related mistake is assuming the list you're reading right now — this one included — will still be accurate in twelve months. It won't be, not entirely. Treat this as a snapshot of the state of the category, and confirm current supply sources directly in your own DSP console before committing a season's budget to a specific list of platforms.