Key takeaways
• MoQ is the 2027 successor to both RTMP and HLS. Sub-second latency at broadcast scale + on-demand in a single protocol — the unified streaming pipeline the industry has chased for a decade.
• 11 vendors demoed at NAB Show 2026. Bitmovin, Cloudflare, AWS, CacheFly, Red5, Synamedia, Akamai, Ant Media, Norsk, Oracle, Nomad Media. nanocosmos shipped first commercial deployment at IBC 2025.
• Browser support is real as of March 2026. All major browsers (Chrome, Edge, Safari 17+, Firefox) now support the WebTransport / HTTP/3 plumbing MoQ depends on.
• The unsolved 2026 question is monetisation. SCTE-35 server-guided ad insertion, regional blackouts, DRM — demoed at IETF MOQ interim Feb 2026 but not yet production-ready across vendors.
• Pilot now, ship in 2027. Forward-looking platforms should run a 6–8-week pilot in 2026 to learn the protocol; production deployments at scale are 2027 territory unless you can absorb beta-grade tooling.
Why Fora Soft wrote this playbook
Fora Soft has shipped 50+ video streaming projects since 2005. StreamLayer for interactive sports streaming, WorldCastLive, Mangomolo, Tradecaster, plus the WHIP/WHEP and AV1 migrations we documented separately.
We have followed the IETF MOQ working group since 2023, ran a 6-week pilot on a draft implementation in late 2025, and reviewed the NAB Show 2026 demos. The pieces in this guide come from that pilot plus public references — nanocosmos’s IBC 2025 commercial launch, Bitmovin’s Cloudflare integration, the IETF MOQ interim Feb 2026.
If you operate a streaming platform and you are looking 18–24 months ahead at protocol roadmap, this guide tells you what MoQ is, who is shipping, where it sits vs WHIP/WHEP and LL-HLS, and when to pilot vs wait.
Planning your 2027 streaming architecture?
Send us your current ingest+delivery topology and 2027 latency targets. We will return a 1-page MoQ migration sketch in 48 hours, free.
What MoQ is in 90 seconds
Media over QUIC (MoQ) is an emerging open standard from the IETF MOQ working group. Goal: deliver real-time, sub-second, AND on-demand video over a single protocol, leveraging modern transport (QUIC, HTTP/3, WebTransport). Promise: replace the bifurcation of RTMP-for-ingest + HLS-for-delivery + WebRTC-for-low-latency with one unified pipeline.
Why this matters. Today’s streaming pipelines are a mess of protocols. Encoder pushes RTMP (TCP-based, deprecated by Adobe). Server packages to HLS (10–30 s latency) or LL-HLS (3–7 s) for the broad audience. WHIP/WHEP for interactive sub-second tier. Each protocol has its own ecosystem, ad insertion, DRM. MoQ proposes a single protocol that handles all three.
Status (May 2026). Working-group drafts mature; multiple production-grade implementations across 11 vendors; nanocosmos shipped first commercial deployment IBC 2025; major browsers support the underlying WebTransport plumbing. Not yet IETF RFC; not yet feature-complete (ad insertion + DRM gaps); 2027 is the production-ready year for most operators.
Why HTTP/3 + QUIC + WebTransport changes the math
QUIC. Connection-oriented transport over UDP, with TLS 1.3 baked in, multiple streams over a single connection (no head-of-line blocking), connection migration (handle network changes mid-flow). The transport that finally makes UDP a first-class citizen on the public internet.
HTTP/3. HTTP semantics over QUIC. Faster connection setup (no TCP+TLS handshake), better behaviour over lossy links, persistent across network changes. Standard since 2022; widely deployed by 2026.
WebTransport. Browser API exposing QUIC streams directly to JavaScript. Lets browsers participate in MoQ without WebRTC’s signalling complexity. The piece that finally makes “single protocol for live + on-demand” viable in browsers.
The combination. MoQ uses WebTransport to talk to browsers, native QUIC for server-to-server, and a relay model that scales to broadcast audience. Sub-second latency at broadcast scale, over standard web infrastructure, without specialised servers or peer-to-peer complexity.
Reference architecture
Figure 1. MoQ reference architecture — producer, origin relay, edge relays, subscriber. Live + on-demand share the protocol.
MoQ vs WHIP/WHEP vs LL-HLS vs HESP
| Protocol | Latency | Scale | 2026 status |
|---|---|---|---|
| RTMP → HLS | 15–30 s | Unlimited (CDN) | Legacy, fading |
| LL-HLS | 3–7 s | Unlimited (CDN) | Mature, default for OTT |
| WHIP/WHEP | 200–500 ms | Up to ~1M with SFU mesh | RFC 9725 finalised; production-ready |
| HESP | 0.5–2 s | Unlimited | Vendor-locked (THEO) |
| MoQ | Sub-second + on-demand | Unlimited (CDN-grade relays) | Multi-vendor demos 2026; production 2027 |
Reach for WHIP/WHEP today (2026) when: sub-second is mandatory and you can ship in 2026. RFC 9725 is finalised; production-ready vendors. See our WHIP/WHEP guide.
Reach for LL-HLS today when: 3–7 s latency is fine, audience is multi-millions, CDN cost matters. The 2026 default for premium OTT.
Reach for MoQ in 2026 only when: you can pilot, you can absorb beta tooling, and you want to claim early-mover advantage on the 2027 protocol.
Reach for MoQ in 2027 when: production deployments mature; ad insertion + DRM stabilise; CDN-grade tooling at price parity with HLS. Most operators’ sweet spot.
Production status across 11 vendors at NAB Show 2026
NAB Show 2026 (Las Vegas, April) was the watershed. Eleven vendors demonstrated MoQ implementations interoperating: Ant Media, AWS, Bitmovin, Broadpeak, CacheFly, Cloudflare, Nomad Media, Oracle, Norsk, Synamedia, Red5. Multi-vendor producer-relay-subscriber chains worked.
nanocosmos shipped first commercial deployment at IBC 2025 via the nanoStream platform. Production-grade for low-volume / high-priority streams; not yet at OTT-broadcast scale.
Cloudflare integration with Bitmovin. Cloudflare provides QUIC infrastructure + edge relays; Bitmovin provides encoder + player. Demonstration deployment at Cloudflare scale; SLA tooling not yet at HLS parity.
AWS. Demoed but not yet a generally available service in 2026. Roadmap signal that AWS Elemental is committing to MoQ for 2027 GA.
Open-source. moq-rs (Rust + TypeScript reference implementation), Eyevinn’s open MoQ library, FFmpeg patches in development. Pilot-grade today.
Browser support reality (March 2026)
Chrome / Edge. WebTransport stable since 2023. MoQ-specific JavaScript libraries available (open-source) for player implementation.
Safari 17+. WebTransport stable. iOS 17+ and macOS Sonoma+ supported.
Firefox. WebTransport stable since version 122 (early 2025).
Smart TV / OTT devices. Connected-TV browsers vary widely. Roku, Fire TV legacy boxes generally do NOT support WebTransport in 2026; modern Tizen/WebOS panels (2024+) do. Plan a fallback to LL-HLS for the long tail of CTV.
Migration playbook from RTMP/HLS
Phase 0 (now, 2026): pilot. 6–8 weeks. Stand up MoQ relay on Cloudflare + Bitmovin or self-hosted moq-rs. Produce one channel through MoQ in parallel with existing RTMP/HLS. Measure latency, quality, vendor maturity. Build internal expertise.
Phase 1 (early 2027): premium tier. Migrate one premium product surface (interactive sports, live commerce, telehealth) where sub-second matters. Keep RTMP/HLS for the cost-sensitive long tail.
Phase 2 (mid-2027): unified pipeline. Consolidate live + on-demand on MoQ. Decommission separate HLS packaging. CDN bills consolidate. Monetisation tooling (SCTE-35 ad insertion, DRM) should stabilise by then.
Phase 3 (late 2027 / 2028): full cutover. RTMP retired. HLS retained only for legacy CTV devices. Multi-codec ladder (AV1 + HEVC + H.264) inside MoQ tracks (see our AV1 guide).
When to wait, when to pilot, when to ship
Wait if: you ship to mass-market OTT today on LL-HLS at 3–5 s, your audience does not need sub-second, your engineering bandwidth is committed elsewhere. MoQ in 2027–2028 will be cheaper and lower-risk.
Pilot if: you have an interactive use case (sports betting, live commerce, telehealth), you want early-mover protocol expertise, you can absorb 6–8 weeks of pilot work in 2026. Pilot lets you learn the protocol shape and vendor maturity without committing production.
Ship now if: you operate a niche premium tier (live betting, financial trading, telehealth specialty), you can absorb beta-grade tooling, you have direct relationships with one of the 11 vendors. Production deployments in 2026 are real but require absorbing operational complexity.
Want to run a 6-8 week MoQ pilot in 2026?
Send us your premium tier (sports / commerce / telehealth) and target latency. We will scope a pilot architecture in 48 hours, free.
Open questions in 2026
1. Server-Guided Ad Insertion (SCTE-35). Demoed at IETF MOQ interim Feb 2026 (Synamedia + Akamai). Not yet a multi-vendor interoperable spec. Production ad insertion at OTT scale is the gating issue for monetisable MoQ deployments.
2. Regional blackouts. Sports rights territoriality. SCTE-35-style markers + relay-side enforcement is the model; needs spec finalisation and multi-vendor test.
3. DRM (Widevine, FairPlay, PlayReady). Premium content needs DRM. MoQ needs hooks for the major DRM systems; Apple FairPlay specifically requires HLS-native packaging that does not yet exist on MoQ.
4. DVR / time-shift unification. The big promise of “same protocol for live and VOD” depends on DVR working seamlessly. Demos exist; production-grade is 2027 territory.
5. CDN economics. Cloudflare claims MoQ via QUIC delivers at HLS-equivalent unit cost. Other CDNs not yet at parity. Will MoQ be cheaper than HLS at scale? The 2027 question.
A decision framework — pick MoQ in five questions
Q1. What is the latency requirement? <1 s mandatory: WHIP/WHEP today, MoQ in 2027. 3–7 s OK: LL-HLS today, MoQ when convenient. >15 s OK: stay on RTMP/HLS until forced.
Q2. Audience type? Niche premium with sub-second need: pilot now. Mass-market OTT: wait. Mixed: deploy LL-HLS for mass + WHIP/WHEP for premium today; consolidate on MoQ in 2027.
Q3. Engineering bandwidth? Yes, 6–8 weeks pilot capacity: pilot now. No: wait for production-grade tooling.
Q4. Monetisation requirements? Need SCTE-35 ad insertion or DRM today: wait for MoQ tooling to mature. Free or subscription tier: pilot is feasible.
Q5. Strategic appetite? Want to be at the front of the curve: pilot 2026, ship 2027, claim early-mover positioning. Risk-averse: wait until 2027–2028 production-grade.
Pitfalls to avoid
1. Migrating production to MoQ in 2026. Tooling is not yet production-grade for OTT-scale. Stay on LL-HLS for production; pilot MoQ in parallel.
2. Ignoring CTV reality. Smart-TV browsers do not all support WebTransport. Always plan an LL-HLS fallback for the CTV long tail.
3. Single-vendor MoQ deployment. Specs are mature enough that interop should work. Build for multi-vendor relay; do not lock to one vendor’s extensions.
4. Skipping the WHIP/WHEP step. Many operators benefit from WHIP/WHEP today (sub-second, RFC 9725 finalised) even if they will move to MoQ in 2027. Do not skip the intermediate; the operational learning carries forward.
5. Underestimating monetisation gaps. Ad insertion + DRM are the production blockers. Pilot is fine without them; production OTT is not.
KPIs to measure
Quality KPIs. Glass-to-glass latency p50/p95 (target: <500 ms / <1 s). First-frame time on viewer join (target: <800 ms). Stall rate (target: <0.5 %). Multi-codec ladder support.
Business KPIs. Cost per delivered minute vs HLS baseline. Engagement uplift on the sub-second tier. Premium-tier conversion rate.
Reliability KPIs. Multi-vendor interop success rate (target: 99 %+). Subscriber join success (target: 99.5 %+).
FAQ
Is MoQ an IETF RFC yet?
Not as of May 2026. Several drafts mature in the IETF MOQ working group; expected RFC publication 2027. Vendor implementations are working against late-stage drafts and interoperating successfully.
MoQ vs WHIP/WHEP — which should I ship?
Today (2026): WHIP/WHEP. RFC 9725 finalised, production-ready, OBS Studio support. 2027+: MoQ when DRM and ad insertion mature. WHIP/WHEP investment carries forward — the operational discipline (low-latency monitoring, sub-second SLAs) is identical.
Will MoQ replace HLS?
Long term, probably yes — for premium tiers. HLS will persist for legacy CTV and cost-sensitive long-tail well into the 2030s. The transition will look like 2010’s RTMP→HLS migration: gradual, with multi-protocol coexistence for years.
What does a pilot cost?
6–8 weeks of senior eng work + Cloudflare/Bitmovin or self-hosted infrastructure. Typical $35–75k all-in for a single-channel pilot. Includes architecture, deployment, latency measurement, vendor evaluation report.
Does MoQ work over poor networks (4G, 5G)?
QUIC handles network changes (4G→5G→Wi-Fi handoff) gracefully — this is one of QUIC’s headline features. Loss recovery is also better than TCP. Early benchmarks suggest MoQ over 4G outperforms RTMP/HLS over 4G; production data still emerging.
Open-source MoQ implementations?
moq-rs (Rust + TypeScript) is the de-facto reference. Eyevinn open-source toolkit is also high-quality. FFmpeg patches in development. All Apache or similar permissive licences. Pilot-grade today; production-grade by 2027.
What about iOS / Apple FairPlay?
Apple FairPlay DRM is HLS-native and does not yet have a clear MoQ integration path. Premium content for iOS in 2026 is HLS-native. MoQ adoption for premium iOS content depends on Apple’s timeline for FairPlay-over-MoQ — not announced as of mid-2026.
Is MoQ better than WebRTC for live?
For broadcast scale (1k+ concurrent), MoQ is architecturally cleaner — relay-based fan-out vs WebRTC SFU complexity. For interactive (1:1, 1:few), WebRTC remains better. The two coexist; MoQ is not replacing WebRTC for video calls.
What to Read Next
Streaming
WHIP/WHEP Replace RTMP
The 2026 sub-second protocol you can ship today.
Codec
AV1 in Production
Pair MoQ delivery with AV1 codec for max bandwidth savings.
Scale
Scale to 1M Viewers
Multi-region relay topology applies to MoQ.
Sports
Interactive Sports Streaming
Sub-second use case — MoQ pilot candidate.
Architecture
Build vs Buy Video SDK
Vendor decisions for streaming infra.
Ready for the 2027 streaming protocol?
MoQ is real. NAB Show 2026 demonstrated 11 vendors interoperating. Production deployments exist (nanocosmos IBC 2025). Browser support is universal. The promise — sub-second + on-demand in one protocol — is on the verge of being claimed.
For most operators, 2027 is the production-ready year. 2026 is the pilot year. Sub-second use cases that need it now: WHIP/WHEP today, migrate to MoQ in 2027 when monetisation tooling matures. The strategic move is to claim the protocol expertise early so the migration in 2027 is a configuration change, not a re-platform.
Want a 6-8 week MoQ pilot scoped for your stack?
Send us your existing topology and target tier. We will return architecture, vendor pick, and 8-week pilot plan in 48 hours, free.
.avif)
Comments