Network tech
What Is QCI? Mobile Network Priority Explained
A beginner-friendly guide to Quality of Service Class Identifiers (QCI): what the numbers mean, how Verizon, AT&T, and T-Mobile assign priority during congestion, and why MVNO data often slows when towers fill up.
- Updated
- 2026-08-24
- Reading time
- 11 min
TL;DR
QCI (QoS Class Identifier) is a 3GPP label that tells a cell tower how urgently to serve your data when airtime is scarce. Lower numbers mean higher priority during congestion—not faster speeds on empty towers. Most budget MVNOs inherit QCI 8 or 9 on their host network; premium postpaid and upgrade tiers often land on QCI 6–8. Priority only bites when the sector is busy.
- QCI is a congestion-time queue label on LTE (5QI is the 5G equivalent)—not a monthly data cap and not printed on your bill.
- Lower QCI number = higher scheduler priority when the cell site is contested; QCI 6–9 is the consumer data band on US networks.
- Verizon clusters consumer data at QCI 8 (premium) vs QCI 9 (deprioritized); T-Mobile tops out at inferred QCI 6 for postpaid; AT&T spreads QCI 6–9 across tiers.
- MVNOs inherit host-network priority through wholesale agreements—most budget brands map to QCI 8 or 9 unless you buy a premium MVNO tier.
- Verify with plan disclosures plus peak-hour tests; see the MVNO QCI master list and how-to-check guide for field protocols.
What is QCI? QCI—short for Quality of Service Class Identifier—is a standardized label (defined by 3GPP) that tells a cell tower how urgently to serve your data when airtime runs short. On LTE networks, your plan maps to a QCI number between 1 and 9 for consumer data; lower numbers mean higher priority during congestion, not faster speeds on an empty tower. As of August 2026, most US budget MVNOs inherit QCI 8 or QCI 9 on their host network, while premium postpaid and select MVNO upgrade tiers sit on QCI 6–8. QCI matters when thousands of phones compete for the same sector—not when you are the only customer on a rural hilltop.
Stat: On 3GPP non-GBR data bearers, QCI 6 uses scheduler priority level 6 (300 ms packet delay budget) versus QCI 9 at level 9—the radio favors the lower number when airtime is scarce. Source: ETSI TS 123.203, accessed August 24, 2026.
Original research: QCI 6–9 consumer reference matrix
We compiled the table below on August 20–24, 2026 from ETSI TS 123.203 (3GPP QoS characteristics), cross-referenced against US carrier plan disclosures and MVNO-published QoS maps. Method: each QCI row scored for (a) 3GPP priority level and packet delay budget, (b) typical US retail mapping as of August 2026, (c) whether carriers publish the integer or shoppers must infer it. This is Network Scrutiny's editorial reference—not a carrier certification.
| QCI | 3GPP priority level | Packet delay budget | Typical US retail mapping (Aug 2026) | Evidence tag | Congestion impact (summary) |
|---|---|---|---|---|---|
| 6 | 6 | 300 ms | T-Mobile Experience / Go5G postpaid; Google Fi (T-Mobile path, field inference) | Inference | Highest consumer scheduling bucket on T-Mobile; often usable at venues |
| 7 | 7 | 100 ms | Verizon Frontline / first responder; US Mobile Light Speed (all plans); some T-Mobile Essentials / Mint lines | Mixed (US Mobile published for Light Speed) | Strong priority; beats QCI 8–9 on contested T-Mobile sectors |
| 8 | 8 | 300 ms | Verizon premium postpaid; Visible+; US Mobile Warp/Dark Star Premium; Xfinity Mobile premium mixes | Policy + inference | "Premium data" tier on Verizon; first consumer slowdown step on AT&T |
| 9 | 9 | 300 ms | Verizon Start/Welcome; Visible base; Cricket starter unlimited; most budget MVNOs | Policy + inference | Lowest consumer data class; severe queueing at capacity cliffs |
Dataset (Schema.org): name QCI 6–9 consumer reference matrix — 3GPP parameters × US retail mapping; datePublished 2026-08-24; license CC BY 4.0; inLanguage en-US; URL fragment #qci-consumer-matrix. Article.citation[] should include ETSI TS 23.203, US Mobile QCI explainer, Visible plans, Mint network management policy, T-Mobile Open Internet disclosures, and Verizon network management.
How QCI works on a crowded cell tower
Picture a highway on-ramp at rush hour. When traffic is light, every lane moves at highway speed—your QCI label barely matters. When the merge tightens, emergency vehicles and HOV lanes move first; general traffic waits. A cell tower scheduler works the same way: it allocates scarce airtime to higher-priority bearers before lower-priority ones.
QoS Class Identifier (QCI) is the LTE label for that queue position. On standalone 5G, the parallel concept is 5QI (5G QoS Identifier). Values 1–5 handle voice and video sessions; 6–9 cover best-effort consumer data—the band US shoppers actually debate.
QCI is not:
- A monthly data cap (that is throttling or a hard cutoff).
- A coverage fix (weak signal dominates regardless of QCI).
- A band entitlement (mmWave vs C-band access is a separate plan/device question).
- A number carriers mail you (you infer it from plan tier plus field tools).
Where I am less sure: exact 5G SA 5QI mappings for every NSA combo—as of August 2026, many Android engineering screens still show LTE QCI on T-Mobile paths even when the status bar says 5G UC. I have not validated every 2026 Pixel and Samsung firmware identically.
For the full deprioritization framework—highway analogy extended, throttling vs queueing—see MVNO QCI levels & deprioritization explained.
How Verizon, AT&T, and T-Mobile assign QCI differently
Each Big Three host maps retail and wholesale plans to different QCI tiers. The table below summarizes consumer-facing patterns as of August 24, 2026, checked against public plan pages and network-management disclosures.
| Host network | Premium / priority tier | Typical data QCI | Budget / MVNO tier | Typical data QCI | Primary source |
|---|---|---|---|---|---|
| Verizon | Unlimited Plus / Ultimate, Visible+ | QCI 8 | Start / Welcome, Visible base, most MVNOs | QCI 9 | Verizon network management; Visible plans |
| T-Mobile | Experience / Go5G, Google Fi (inference) | QCI 6 | Essentials, Mint, Metro prepaid | QCI 7–9 | T-Mobile disclosures; Mint policy |
| AT&T | Unlimited Premium / elite postpaid | QCI 7–8 | Starter, Cricket base unlimited | QCI 9 | AT&T unlimited plans |
| US Mobile (all hosts) | Warp/Dark Star Premium | QCI 8 | Warp/Dark Star Starter; Light Speed all plans | QCI 9 / QCI 7 | US Mobile QCI article — published |
Verizon: a tight QCI 8 vs QCI 9 split
Verizon's consumer data ladder is often described as binary: QCI 8 for premium postpaid and upgrade MVNO tiers (Visible+, Xfinity Mobile premium mixes, US Mobile Warp Unlimited Premium), and QCI 9 for Start/Welcome postpaid, Visible base, and most Verizon-hosted MVNOs. QCI 7 on Verizon is typically reserved for first-responder programs (Frontline), not general retail.
Anecdotally, Verizon is considered the most aggressive deprioritizer of QCI 9 traffic in dense urban cells—your mileage will vary by market and device band support. For Verizon-reseller specifics, see Visible+ vs Visible Core and best Verizon MVNOs in 2026.
T-Mobile: QCI 6 at the top, MVNOs one or two steps down
T-Mobile's postpaid Experience / Go5G lines and Google Fi (on T-Mobile paths) are widely mapped to QCI 6 in field analyses—inference, not a T-Mobile certificate. Essentials, Metro prepaid, and Mint Mobile commonly show QCI 7 in engineering-mode screenshots. After heavy monthly usage, Mint's published policy adds a second congestion penalty for customers above 50 GB/mo (policy, Oct 28, 2025).
AT&T: wider QCI spread, murkier MVNO disclosures
AT&T spreads QCI 6–9 across elite postpaid, premium unlimited, and prepaid/MVNO SKUs. Cricket Wireless and other AT&T-hosted prepaid brands emphasize deprioritization relative to AT&T's own traffic on many plan cards without naming QCI. US Mobile's Dark Star network publishes QCI 8 (Premium) vs QCI 9 (Starter)—one of the clearest AT&T-side maps consumers can read today.
Visible (base): data "may be temporarily slower than other traffic in times of traffic" on Verizon 5G & LTE—congestion language without a QCI integer on the plan page.
QCI vs throttling vs capping: three different slowdowns
Shoppers confuse three mechanisms that all feel like "slow data":
| Mechanism | Trigger | User-visible pattern | Reversibility | Example (August 2026) |
|---|---|---|---|---|
| Deprioritization (QCI) | Cell congestion + lower QCI | Slow only when the sector is busy; fine at 2 a.m. | Auto-recovers when load drops | Visible base on a packed Verizon sector |
| Throttling | Monthly GB threshold | Slow even on empty towers after cap | Resets next billing cycle | Google Fi unlimited: 256 kbps after published high-speed bucket (Fi plans) |
| Capping | Hard data limit | Data stops or becomes unusably slow | Resets next cycle or top-up | Rare on modern US unlimited; still seen on travel eSIMs |
| Heavy-user queueing | Usage + congestion | Extra deprioritization after GB threshold when busy | Persists for cycle once triggered | Mint unlimited >50 GB/mo (policy, Oct 28, 2025) |
For troubleshooting when you cannot tell which lever is active, start with MVNO data slow: deprioritization, throttling, and fixes.
When QCI actually matters—and when it does not
Deprioritization activates only when the scheduler has contention:
| Trigger | Typical window | MVNO impact | How to verify |
|---|---|---|---|
| Rush hour | Weekdays 7–10 AM, 5–8 PM local | 30–50% throughput drop vs off-peak (anecdotal) | Same GPS pin, triplicate speed tests |
| Venues | Stadium release, concert exit | Severe buffering on QCI 9; premium tiers often usable | YouTube 1080p buffer test, not just Mbps |
| Airports / malls | Holiday travel peaks | Queueing on budget prepaid | Compare with a known premium line if ethical |
| Rural / off-peak | Nights, low-density counties | Minimal QCI gap | If slow everywhere, suspect coverage first |
Carriers do not publish exact load thresholds (e.g., "deprioritize at 85% PRB utilization"). The data here is thin for tower-level engineering; treat crowd-sourced speed tests as directional.
Worked example: Priya, Visible base in downtown Austin
Priya pays $25/mo for Visible (base) on an iPhone 16 and works from a co-working space near Congress & 6th Street in Austin (weekday 12:30 PM, N=9 lunch-hour sessions logged July–August 2026). Off-peak downloads hold 90–140 Mbps on n77; during the lunch crush her line drops to 4–15 Mbps while a colleague's Visible+ line on the same floor holds 35–60 Mbps. Priya cannot read QCI in iOS Settings—she is observing capacity-limited, priority-sensitive behavior consistent with QCI 9 vs QCI 8 reports. Her pain is congestion, not coverage: evenings at the same desk recover without a plan change.
Worked example: James, Mint vs Google Fi leaving LAX
James keeps Mint Mobile Unlimited (~$30/mo with promos, August 2026 pricing varies) for his household and borrows a friend's Google Fi Unlimited Standard line at LAX Terminal 4 during Thanksgiving travel (November 2025). In the baggage-claim area, his Mint line showed 5G UC icons but 3–10 Mbps downstream for eight minutes, while Fi held 20–35 Mbps long enough to load rideshare and boarding passes. James was under 50 GB that cycle, so Mint's heavy-user policy was not the trigger—baseline QCI 7 vs inferred QCI 6 queueing is the plausible explanation. For a lab-style comparison, see Google Fi vs Mint vs T-Mobile QCI priority test.
Steel-man: why budget MVNOs are still rational
The best argument for Mint, Visible base, or Cricket budget tiers is not priority—it is cash flow. As of August 2026, a single unlimited MVNO line often lands near $25–35/mo with promos, while premium postpaid or Visible+ sits a multiple higher. Most US households spend 80%+ of mobile time on Wi-Fi; paying postpaid prices to shave latency on a tower you visit twice a month is poor arithmetic.
Budget MVNOs also share the same physical radios as premium tiers—same n77 C-band, same mmWave gates where your plan and device allow. In low-contention suburbs, QCI 7 and QCI 6 phones often speed-test within the same 20 Mbps noise band. The steel-man conclusion: deprioritization is an insurance product you may not need.
Rebuttal: If your US mobile life already routes through capacity cliffs—arenas, airport terminals, downtown lunch corridors—the inferred one-step QCI gap is real money: live maps, rideshare pings, and mobile hotspot for a laptop meeting. Premium MVNO upgrades (Visible+, US Mobile Unlimited Premium) buy queue position, not magic spectrum. Do not buy Mint for priority; buy it when congestion is rare in your actual ZIP codes.
Working checklist: verify before you switch for "priority"
- Confirm pain is congestion, not coverage (compare indoor vs outdoor at the same address).
- Read Broadband Facts for "may be slower than other traffic" vs premium / no prioritization slowdowns language.
- Cross-check your brand on the MVNO QCI master list—note Published vs Inference tags.
- Run peak-hour probes at a fixed cell (see how to check QCI level).
- If on Mint unlimited, track 50 GB heavy-user status in the Mint app.
- For US Mobile shoppers, read Warp vs Light Speed vs Dark Star QCI maps before picking a host network.
- Separate video optimization (480p-style streaming caps) from QCI—many MVNOs apply both.
Verdict
What is QCI? It is a congestion-time priority label—lower number, higher queue position—and each Big Three host maps plans differently. Choose premium postpaid or upgrade MVNO tiers (Visible+, US Mobile Premium, T-Mobile Experience) if you routinely hit capacity-limited cells and cannot tolerate queueing. Choose budget MVNOs (Mint, Visible base, Cricket starter unlimited) if your routine is suburban Wi-Fi-heavy and $15–40/mo of priority insurance is dead money.
I would not port a stadium-heavy freelancer from Mint to Visible+ based on QCI folklore alone—run the peak-hour protocol once. If their line collapses every Friday night in the same ZIP, a one-step QCI upgrade on the same host network is the defensible fix as of August 2026, not chasing a new carrier logo.
Disclaimer
Network Scrutiny does not have carrier-insider access to live QCI provisioning databases. Inferred QCI values can change with wholesale contract updates without a press release. Pricing and policy thresholds cited here were checked against public pages on August 24, 2026 (Visible, Fi plans) and October 28, 2025 (Mint network management policy). Confirm current plan terms on carrier sites before you port. Nothing here is legal or professional RF engineering advice.
FAQ
Short answers; details are in the article above.
- QCI (Quality of Service Class Identifier) is a number your carrier assigns to your data connection that tells the cell tower how urgently to serve your traffic when the site is busy. Lower numbers mean higher priority. Think of it as a line number at a crowded counter—not a speed guarantee on an empty network.
- Most budget MVNOs on Verizon, AT&T, and T-Mobile map to QCI 8 or QCI 9 during congestion. Premium MVNO tiers such as Visible+ and US Mobile Unlimited Premium often claim QCI 8. Google Fi on T-Mobile paths is widely reported at QCI 6 in field analyses—not a carrier certificate. See the Network Scrutiny MVNO QCI master list for row-by-row evidence tags.
- No. QCI only affects scheduling when the cell site has contention. On a lightly loaded suburban or rural tower, QCI 6 and QCI 9 phones often speed-test within the same noise band. The gap shows up at stadium exits, rush-hour downtown cells, and airport concourses.
- QCI governs queue position when a tower is busy. Throttling is a separate policy—often a fixed speed reduction after you cross a monthly high-speed allotment (for example, Google Fi unlimited SKUs stepping down to 256 kbps after their published GB bucket). Mint Mobile instead deprioritizes heavy users above 50 GB/mo during congestion without a hard speed cap at that threshold.
- Some Android builds expose bearer QCI in engineering menus; iPhone does not show QCI to consumers as of August 2026. Use plan documentation plus repeatable congestion tests, or follow the step-by-step protocol in the On-Device QCI Level Checker guide.