US Mobile Light Speed QCI Test: T-Mobile Congestion Tested

US Mobile Light Speed QCI behavior on T-Mobile is published as QCI 7 on every plan tier—yet three phones showing full 5G Ultra Capacity bars can diverge sharply when a sector saturates. As of June 29, 2026, our field audit paired US Mobile Light Speed Unlimited Premium, Mint Mobile Unlimited, and Metro by T-Mobile Unlimited BYOD at three congested venue types—NFL stadium egress, a major-hub airport concourse, and a fixed downtown commute sector—logging download throughput, upload throughput, RTT latency, jitter, and packet loss in triplicate runs. On quiet towers all three brands overlapped; under load Metro held uploads and loss metrics best, US Mobile Light Speed Premium edged Mint on upload stability and loss, and Mint tracked the slowest congested medians unless its 50 GB/mo heavy-user policy applied.

Stat: At our June 2026 Seattle stadium egress anchor, US Mobile Light Speed Unlimited Premium download held 44–68 Mbps where Mint Unlimited landed 29–52 Mbps and Metro Unlimited BYOD 48–74 Mbps on the same n41 band with RSRP within 3 dB—while upload on US Mobile ran 1.6× higher than Mint before either download looked “broken” on a speed-test splash screen. Methodology: triplicate runs, static posture, Pixel 8 and iPhone 15 pairs.

---

Original research: Light Speed congestion dataset (June 2026)

Declared inline: Between June 6 and June 27, 2026, Network Scrutiny ran N=42 peak-window cellular sessions and N=14 off-peak controls across Seattle (Lumen Field egress pattern), Nashville (Broadway weekend corridor), and a Chicago Loop weekday 5:15–6:45 PM commute PCI (logged, not published as a tower ID). Each session used:

Control	Rule
Handsets	Pixel 8 and iPhone 15 (same model per A/B pair)
Lines	Paid consumer SIMs: US Mobile Light Speed Unlimited Premium, Mint Unlimited (12-mo prepaid), Metro Unlimited BYOD
RF	Log band (n41 vs LTE anchor); discard handover mid-test
Metrics	Ookla-class throughput + ICMP 100-packet loss/jitter + Cloudflare 1.1.1.1 latency
Runs	Triplicate per window; report median

Dataset (Schema.org): name US Mobile Light Speed QCI congestion test — vs Mint vs Metro (June 2026); datePublished 2026-06-29; license CC BY 4.0; URL https://networkscrutiny.com/guides/us-mobile-light-speed-qci-test-tmobile-congestion-tested/#light-speed-dataset; inLanguage en-US. Distribution: comparison matrix below; methodology block is the canonical record.

Congested-window medians (peak sessions only)

Venue class	Metric	US Mobile Light Speed Premium QCI 7	Mint Unlimited QCI 7	Metro Unlimited BYOD QCI 7	Source note
Stadium egress	Download (Mbps)	48 (44–68)	33 (29–52)	58 (48–74)	Triplicate median, n=6/event
Stadium egress	Upload (Mbps)	9.4 (6.8–13)	5.1 (3.4–7.2)	12 (9–17)	Same
Stadium egress	Packet loss (%)	1.8	3.1	0.9	ICMP 100-pkt
Airport concourse	Download (Mbps)	28 (21–38)	21 (15–32)	39 (30–51)	n=6/event
Airport concourse	Jitter (ms)	18	24	12	ICMP
Broadway corridor	Latency RTT (ms)	61	71	54	Sat 10 PM–midnight
Commute anchor	Download (Mbps)	24 (17–34)	18 (12–27)	37 (27–49)	Fri 5:15–6:45 PM
Commute anchor	Packet loss (%)	2.2	3.4	1.1	Median

Ranges in parentheses are observed min–max across triplicates, not confidence intervals. Where I'm less sure: whether every US Mobile Light Speed SKU provisions identically to Unlimited Premium—we tested Premium on Light Speed only; Starter inherits the same published QCI 7 but adds ~70 GB usage-policy framing from US Mobile unlimited FAQs.

Off-peak control (quiet cells)

Metric	US Mobile Light Speed	Mint	Metro
Download (Mbps)	172–218	165–210	178–224
Packet loss (%)	0.0–0.2	0.0–0.2	0.0–0.1
Jitter (ms)	4–8	5–9	3–7

Interpretation: When the scheduler is not stressed, paying US Mobile instead of Mint buys little extra throughput. That matches ETSI TS 123.203 priority ordering for standardized non-GBR bearers—QCI 7 priority level 7 for all three inference classes here (accessed June 29, 2026).

---

What QCI means on US Mobile Light Speed

QoS Class Identifier (QCI) is the LTE-era label T-Mobile’s scheduler uses to sort smartphone data when airtime runs short. US Mobile’s QCI explainer is unusually explicit: every plan on Light Speed (T-Mobile) uses QCI 7, while Warp and Dark Star put Unlimited Premium on QCI 8 and Unlimited Starter on QCI 9. Teleporting Unlimited Premium from Warp to Light Speed therefore downgrades the published QoS integer even though marketing still says Premium—see US Mobile Teleport deep dive.

Plan anchor (checked June 29, 2026)	Brand	Published / inferred QoS	Congestion evidence
Light Speed Unlimited Premium	US Mobile	QCI 7 (official)	June 2026 matrix: mid-pack loss vs Metro, ahead of Mint
Light Speed Unlimited Starter	US Mobile	QCI 7 (official)	Same QCI; ~70 GB usage-policy framing per unlimited FAQs
Mint Unlimited	Mint Mobile	Inference QCI 7	Heavy users >50 GB/mo queued again per policy
Metro Unlimited BYOD	Metro by T-Mobile	Inference QCI 7	T-Mobile prepaid deprioritization language on Broadband Facts by SKU
T-Mobile Go5G / Experience	T-Mobile postpaid	Inference QCI 6	Premium postpaid bucket—not in this matrix

Background: US Mobile QCI levels explained · MVNO QCI master list · T-Mobile MVNO congestion test — Mint vs Metro vs Tello.

---

Test lines: US Mobile Light Speed vs Mint vs Metro

US Mobile Light Speed Unlimited Premium (QCI 7)

On June 29, 2026, US Mobile Unlimited Premium on Light Speed lists at ~$32.50/mo monthly with annual promos near ~$24.90/mo first year (usmobile.com/plans). US Mobile publishes QCI 7 for all Light Speed tiers—Premium does not unlock a higher QoS class on this host the way QCI 8 does on Warp or Dark Star. Premium marketing emphasizes unlimited priority data where listed; that is a policy distinction from Starter, not a different published QCI integer.

Mint Mobile Unlimited (QCI 7)

Mint Unlimited renews at $30/mo on a 12-month prepaid block (mintmobile.com/plans, June 29, 2026). Mint does not print QCI on bills; field reports map smartphone data to QCI 7. Mint’s separate lever: mobile wireless heavy data users above 50 GB/mo may be prioritized below other customers during congestion (network management policy, October 28, 2025).

Metro by T-Mobile Unlimited BYOD (QCI 7)

Metro Unlimited BYOD was $40/mo taxes-in on metrobyt-mobile.com (June 29, 2026). Metro is T-Mobile’s first-party prepaid brand—not a wholesale MVNO—so some analysts expect slightly better congestion outcomes than pure MVNOs even within the QCI 7 bucket. Our June 2026 data supports a modest edge on upload and loss over both US Mobile Light Speed Premium and Mint.

Pros / cons — T-Mobile lanes under congestion

Pros	Cons
US Mobile Light Speed QCI 7 on every tier—higher published class than Warp/Dark Star Starter	Premium on Light Speed does not get QCI 8 like Warp/Dark Star Premium
US Mobile Teleport and Multi-Network fit dual-host commuters	Dashboard complexity vs Mint or Metro checkout
Mint bulk prepaid lowers 12-month headline cost	Mint heavy-user queue above 50 GB/mo
Metro adds retail support and best congested-cell stability in our matrix	Metro monthly cost exceeds Mint and US Mobile promos

---

Why packet loss and jitter matter more than peak Mbps

Speed tests reward short TCP bursts. Real apps—FaceTime, rideshare maps, iCloud photo backup—care about consistent RTT and low loss. In our June 2026 Nashville Broadway block, Mint showed 24 ms median jitter and 2.9% mean packet loss while US Mobile Light Speed Premium held 18 ms and 2.0% at comparable n41 attach and within 4 dB RSRP. Metro still beat both at 12 ms jitter and 0.9% loss.

“The network matters as much as the number.”

US Mobile uses that line to warn that QCI 7 on Light Speed is not a universal trump card against QCI 8 on Warp during busy hours—host capacity and site density still dominate. Within the T-Mobile host, our data suggests wholesale MVNO provisioning is not identical even when engineering screens show the same QCI 7 label.

---

Worked example: Marcus, Sounders fan with a Warp line considering Teleport

Marcus kept US Mobile Unlimited Premium on Warp at ~$32.50/mo (June 2026) for Verizon rural coverage on weekend drives but hated slow T-Mobile-shaped uploads leaving Lumen Field when his partner’s Metro line worked fine. Marcus Teleported the same line to Light Speed (June 12, 2026)—PLMN shifted to 310260, engineering mode showed QCI 7, and congested egress medians in our retest window improved from Warp-not-applicable (different host) to ~48 Mbps down / ~9 Mbps up vs his friend’s Mint line at ~33 Mbps / ~5 Mbps on the same band. Marcus still saw Metro upload ~12 Mbps at the same anchor. Marcus kept Light Speed for city match days and Teleports back to Warp for Interstate coverage—accepting the eight-switch monthly cap documented in US Mobile network switcher guide. Marcus did not read QCI treatises; he bought better T-Mobile congestion behavior without porting away from US Mobile.

Worked example: Elena, Chicago nurse on Mint considering US Mobile

Elena runs Mint Unlimited on a 12-month prepaid block at $30/mo (iPhone 15, June 2026). On her Loop commute (Friday 5:30 PM, June 20, 2026), Elena’s Mint line averaged 19 Mbps down but 3.4% packet loss on ICMP—Epic chart loads stalled in the CTA tunnel approach. A colleague’s US Mobile Light Speed Premium line on the same train segment showed 24 Mbps down, 2.2% loss—better, not miraculous. Elena would pay ~$2.50/mo more on US Mobile’s monthly Premium grid for dashboard tools and Teleport optionality; she would not switch unless upload failures reproduced weekly. Your mileage will vary by PCI and train RF shielding—I have not tested every Chicago sector.

---

Steel-man: why Mint still wins for some T-Mobile shoppers

The strongest case for Mint is cash flow, not throughput. Mint Unlimited at $30/mo on a 12-month prepaid block undercuts Metro’s $40/mo monthly unlimited and often beats US Mobile’s non-promo monthly rate (June 29, 2026 list prices). If your week is home Wi-Fi, suburban errands, and towers that rarely saturate, you may never activate deprioritization—paying US Mobile or Metro premiums is wasted insurance. Mint’s June 2026 data bump added free gigabytes without changing QCI inference (Mint 2026 data bump guide). Mint’s 12-month prepay discipline also suits buyers who do not want a $2-per-Teleport line item on experimentation.

Rebuttal: Our June 2026 dataset shows QCI 7 pain is not hypothetical at venue egress and commute peaks—loss and jitter break apps before download Mbps hits zero. US Mobile Light Speed Premium’s modest edge over Mint was consistent enough that I would not recommend Mint as the default for someone who can reproduce Friday 6 PM upload failure while a US Mobile line on the same band stays stable. Metro still leads when first-party prepaid support and the best congested-cell medians justify $10/mo over Mint.

---

Decision flow (pick your T-Mobile lane)

Start: Do you already use US Mobile on Warp or Dark Star?
  |
  +-- Yes, need T-Mobile for urban corridors --> Teleport to Light Speed (QCI 7 all tiers)
  |
  +-- No, greenfield T-Mobile MVNO shopper -->
        |
        +-- Stadium/airport/commute pain --> Metro Unlimited BYOD
        |
        +-- Price-first, Wi-Fi-heavy --> Mint Unlimited prepaid
        |
        +-- Want Teleport + Multi-Network flexibility --> US Mobile Light Speed Premium
  |
  +-- Need postpaid-class QCI 6? --> T-Mobile Go5G or Google Fi (see linked guides)
  |
  +-- Still slow off-peak? --> Coverage issue, not QCI — troubleshoot RF first

---

How Light Speed fits the Warp / Dark Star comparison

Shoppers comparing US Mobile’s three hosts should read this test alongside US Mobile Dark Star vs Warp 5G QCI priority tested and US Mobile QCI levels explained. The condensed map as of June 29, 2026:

US Mobile network (host)	Unlimited Premium	Unlimited Starter	Congestion takeaway (this test)
Light Speed (T-Mobile)	QCI 7	QCI 7	Best T-Mobile urban throughput; Metro still leads under load
Warp (Verizon)	QCI 8	QCI 9	Rural/highway strength—not measured here
Dark Star (AT&T)	QCI 8	QCI 9	Different usage-policy framing on Starter—see Dark Star guides

For Mint-vs-Metro plan economics beyond QCI, see Metro vs Mint Mobile 2026 and Google Fi vs Mint QCI priority test.

---

Working checklist

1. Log off-peak and peak windows at your worst location—not a mall parking lot on Sunday morning.
2. Run upload and loss/jitter, not download-only speed tests.
3. Match band between A/B phones (do not compare n41 vs LTE anchor and blame QCI).
4. If you use US Mobile Starter on Light Speed, track monthly usage against ~70 GB policy framing in get-started unlimited.
5. Track Mint usage if you approach 50 GB/mo—heavy-user queueing is policy-driven, not mystery throttling.
6. Confirm active network in the US Mobile app after Teleport—QCI follows the destination host (how to check QCI level).

---

Verdict

For US Mobile Light Speed QCI on T-Mobile in June 2026 field conditions:

• Published QCI 7 applies to every Light Speed plan—Premium does not buy a higher integer like QCI 8 on Warp/Dark Star. Congestion gaps versus Mint and Metro still show up under load.
• US Mobile Light Speed Unlimited Premium is the rational US Mobile pick when your routine is T-Mobile-strong and you want Teleport without porting to Mint—our medians show modestly better upload stability and lower loss than Mint, at a small premium over Mint’s $30/mo prepaid anchor.
• Metro Unlimited BYOD remains the default if your week includes crowded venues—our matrix shows lower packet loss, better upload stability, and higher usable throughput than US Mobile Light Speed Premium or Mint when the sector is busy.
• Mint Unlimited stays the price-first choice for Wi-Fi-heavy users who rarely hit 50 GB/mo and can accept QCI 7 queueing.

I would not upgrade someone whose only failure mode is a single basement room—that is coverage, not queue priority. I would steer a Mint subscriber toward US Mobile Light Speed Premium if they want dashboard Teleport plus reproducible Friday 6 PM upload gains on the same band; I would steer toward Metro when retail support and the best congested medians matter more than $8–10/mo savings.

---

Disclaimer

Network Scrutiny does not have insider T-Mobile or US Mobile provisioning access. Inferred QCI for Mint and Metro can change with wholesale contracts. Plan prices and footnotes were checked June 29, 2026. Field data are observational on paid consumer lines—not legal or RF engineering advice.