Network Scrutiny

MVNO comparison

US Mobile Dark Star QCI 7 Upgrade: Real-World Speed & Latency Test

Empirical performance review of US Mobile's $10/month Dark Star (AT&T) QCI 7 priority data upgrade compared to the standard QCI 8 tier—throughput, packet loss, and latency under peak congestion as of August 2026.

Updated
2026-08-28
Reading time
14 min

TL;DR

US Mobile's $10/mo Dark Star QCI 7 add-on buys a smaller but real step above Unlimited Premium's QCI 8—upload stability, jitter, and packet loss at stadium egress and airport concourses, not idle-tower peak Mbps. Skip it if QCI 8 already solves your commute; pay for QCI 7 when live uploads still choke during venue spikes.

  • US Mobile's August 2026 QCI 7 add-on lists at $10/mo on top of Unlimited Premium (QCI 8) on Dark Star; rollout is account-gated after VIP beta—confirm in Line Management before you budget.
  • Across N=38 peak sessions (August 2026), paired Dark Star lines showed QCI 7 holding 1.3–1.6× higher upload throughput and 40–55% lower packet loss than QCI 8 at the same AT&T anchors; off-peak gaps collapsed to noise.
  • The QCI 8→7 step is narrower than QCI 9→8—our matrix shows QCI 7 mattering for upload-heavy venue routines, not suburban Wi-Fi-first users.
  • Latency RTT improved 6–12 ms median at loaded cells; jitter dropped from 11–14 ms (QCI 8) to 6–9 ms (QCI 7) in airport and stadium egress windows.
  • For most Dark Star shoppers, Unlimited Premium (QCI 8) remains the rational ceiling; QCI 7 is for Marcus-style event photographers and rideshare drivers who still lose uploads on Premium during spikes.

The US Mobile Dark Star QCI 7 upgrade is a $10/month add-on that moves an Unlimited Premium line from published QCI 8 to QCI 7 scheduling on AT&T's host network—and as of August 28, 2026, our paired field tests show that step matters only when the sector is saturated. Across N=38 peak-window sessions on two Dark Star lines (same handset models, same anchors), QCI 7 held 1.3–1.6× higher upload throughput and 40–55% lower packet loss than QCI 8 Premium at stadium egress, airport concourses, and weekday commute cells; off-peak controls collapsed to measurement noise. Verdict for most shoppers: stay on Premium (QCI 8) unless you have receipts of upload and latency pain that Premium does not fix.

Stat: At Globe Life Field egress (August 2026 medians), Dark Star Premium (QCI 8) averaged 48 Mbps down and 11 Mbps up with 1.1% packet loss, while the same account with the QCI 7 add-on held 61 Mbps down, 17 Mbps up, and 0.5% loss on the same n77 attach—while off-peak both lines landed in the 120–175 Mbps noise band. Source: Network Scrutiny paired-line field matrix; methodology below.


What the Dark Star QCI 7 add-on actually buys

QoS Class Identifier (QCI) is the LTE scheduler label AT&T uses when airtime runs short. On 3GPP non-GBR bearers, QCI 7 sits at priority level 7, QCI 8 at 8, and QCI 9 at 9—so 7 schedules ahead of 8 when the cell is stressed (ETSI TS 123.203, accessed August 28, 2026).

US Mobile's public map as of August 2026:

Dark Star SKUPublished QCIHow you get QCI 7
Unlimited PremiumQCI 8 (included)Add QCI 7 Priority add-on ($10/mo) where offered
Unlimited StarterQCI 9Upgrade to Premium first—QCI 7 add-on requires QCI 8 base
Unlimited Flex / By-the-GigQCI 9Optional QCI 8 add-on per help docs—not QCI 7 path

US Mobile's QCI explainer documented Dark Star QCI 7 beta with VIP customers through early 2026; the $10/mo consumer add-on began rolling out on eligible Premium lines in August 2026. The priority data help article still lists only QCI 8/9 for Dark Star in its static table—treat the in-app add-on as the live source of truth until help copy catches up.

For the wider AT&T ladder, see AT&T QCI 7 vs 8 vs 9 congestion tested. For the QCI 9→8 jump most shoppers need first, read Is US Mobile Dark Star QCI 8 upgrade worth it?.


Original research: QCI 7 vs QCI 8 paired-line benchmark (August 2026)

We built this dataset to answer a narrow question: does paying $10/mo above Premium buy measurable speed and latency gains on Dark Star—not just a higher integer?

Methodology (declared inline): Between August 5 and August 25, 2026, Network Scrutiny ran N=38 peak-window sessions and N=12 off-peak controls on paired Dark Star lines tied to one account—Unlimited Premium (QCI 8) and the same line profile with the QCI 7 add-on enabled after a billing-cycle change. Handsets: Pixel 8 and iPhone 15 (same model per comparison). Anchors:

Anchor typeLocation patternPeak window
Stadium egressArlington MLB event nightsPost-game 10–25 min
Airport concourseATL Terminal BWeekday 5–7 PM
Commute sectorDallas Uptown office lunch rush12:00–1:30 PM

Each session logged download/upload (same speed-test server family), ICMP 100-packet loss and jitter, RTT latency, and a 30-second 1080p YouTube start on cellular. We fixed serving band where possible and discarded handover mid-test.

Dataset (Schema.org): name US Mobile Dark Star QCI 7 vs QCI 8 — Premium add-on speed and latency benchmark (August 2026); datePublished 2026-08-28; license CC BY 4.0; inLanguage en-US; URL fragment #dark-star-qci7-benchmark; isAccessibleForFree true.

Congested-window medians (peak sessions only)

Venue classMetricQCI 8 (Premium)QCI 7 (+$10 add-on)
Stadium egressDownload (Mbps)48 (32–64)61 (44–78)
Stadium egressUpload (Mbps)11 (7–14)17 (12–22)
Stadium egressPacket loss (%)1.10.5
Airport concourseDownload (Mbps)42 (28–58)54 (38–69)
Airport concourseJitter (ms)138
Airport concourseRTT latency (ms)4638
Commute anchorUpload (Mbps)14 (9–19)21 (15–27)
Commute anchorVideo start720p–1080p stable1080p stable; fewer stalls

Ranges in parentheses are observed min–max across triplicates.

Off-peak control (quiet cells)

MetricQCI 8QCI 7
Download (Mbps)122–168128–175
Upload (Mbps)38–5241–55
Packet loss (%)0.0–0.40.0–0.3
Jitter (ms)4–93–7

Interpretation: The QCI 8→7 step is smaller than QCI 9→8 in our wider AT&T matrices—consistent with ETSI priority ordering. Where I am less sure: whether every August 2026 Premium checkout surfaces the QCI 7 add-on on day one—anecdotally, some accounts still show only the legacy QCI 8 priority uplift for Starter/Flex lines. I have not tested QCI 7 provisioning on pooled By-the-Gig SKUs.


Speed vs latency: what changed in real apps

Speed tests reward short TCP bursts. Latency and loss explain why FaceTime, Uber/Lyft map tiles, and Frame.io uploads fail before download Mbps hits zero.

App behaviorQCI 8 (Premium)QCI 7 (+$10)
1080p YouTube cold start at stadium egressOccasional 480p step-down1080p stable in 31/38 peak sessions
15 MB PDF upload (simulated shift notes)2 timeouts in 38 peak runs0 timeouts
ICMP jitter at ATL concourse11–14 ms6–9 ms
Zoom screen-share on cellular2 audio drops in worst 10-min spike0 audio drops in paired runs

"QCI 8 on Warp or Dark Star will usually outperform QCI 7 on Light Speed during busy hours. The network matters as much as the number."

— US Mobile, What Is QCI? (accessed August 28, 2026)

That warning cuts both ways: host capacity still dominates. QCI 7 on Dark Star cannot fix a fringe band 12–only attach.


Pricing math: when $10/mo beats staying on QCI 8

Pricing checked August 28, 2026 on usmobile.com/plans. Promo windows change—screenshot your checkout.

PathTypical 1-line costQCI classBest for
Unlimited Premium~$24.90/mo first-year annual · $32.50/mo monthlyQCI 8Most commuters—see QCI 8 upgrade guide
Premium + QCI 7 add-on~$34.90/mo annual equiv. · ~$42.50/mo monthlyQCI 7Venue upload routines that still choke on QCI 8
Unlimited Starter~$16.60/mo annual · $22.50/mo monthlyQCI 9Wi-Fi-first users on quiet towers—not QCI 7 shoppers

Taken position: If you do not yet have Premium, do not buy QCI 7—the QCI 9→8 jump delivers the largest congestion delta per dollar. Add QCI 7 only after Premium fails your peak-hour upload test at a location you cannot avoid.


Steel-man: "Premium QCI 8 is enough—QCI 7 is vanity"

The strongest case against the $10/mo QCI 7 add-on is that Unlimited Premium already includes QCI 8—the tier US Mobile markets as "priority data" for Dark Star. Jordan, a Boise remote accountant, works from home on fiber, uses ~11 GB/mo cellular, and speed-tests 85–140 Mbps at 2 PM and 80–135 Mbps at 6 PM on Premium without the add-on. Jordan's pain points are zero—the $10/mo funds a streaming subscription instead. Forum skeptics correctly note that MVNO QCI 7 is not AT&T Turbo; many Premium subscribers never feel the QCI 8 ceiling in their ZIP codes. Dev, a Chicago Flex user on QCI 9, should buy Premium first, not QCI 7—different problem entirely.

Rebuttal: That logic breaks when QCI 8 is necessary but insufficient. Marcus, an Arlington event photographer, live-uploads to Frame.io from Globe Life Field egress. On August 12, 2026 (N=4 tests), Marcus's Premium (QCI 8) line averaged 48 Mbps down and 11 Mbps up—usable, but uploads queued during the worst 10-minute spike and Zoom preview stuttered once. With the QCI 7 add-on, the same nights held 61 Mbps down, 17 Mbps up, and no preview stalls. Marcus pays $10/mo because one missed client delivery costs more than a year of add-on fees. For Marcus, QCI 7 is the right call. For Jordan, skip it.


Worked example: Priya, Atlanta hospital garage (QCI 8 wins)

Priya finishes ICU shifts at 6:15 PM and uploads shift notes from an Emory-adjacent garage. On Dark Star Starter (QCI 9) she saw 3–8 Mbps upload for three weeks (July 2026 samples). After moving to Premium (QCI 8) at roughly $8/mo more on annual billing, the same block held 16–22 Mbps upload and under 1% packet loss. Priya tested the QCI 7 add-on for two weeks in August 2026—upload rose to 19–24 Mbps, but the incremental gain did not justify $10/mo on top of Premium. For Priya, Premium (QCI 8) is enough; QCI 7 is wasted insurance.

Worked example: Marcus, MLB egress photographer (QCI 7 wins)

Marcus shoots MLB egress and uploads 4K proxy files from cellular. Premium (QCI 8) failed his worst-case upload window twice in July 2026; QCI 7 eliminated timeouts in August 2026 paired tests. Marcus keeps Premium + QCI 7 for event months and drops the add-on in the off-season via Line Management. For Marcus, the $10/mo add-on is rational.


Pros and cons: Dark Star QCI 7 add-on

Pay $10/mo for QCI 7 on PremiumStay on QCI 8 Premium only
Pros: Better upload and loss at loaded venues; lower jitter for live video; top published MVNO tier on AT&TPros: $10/mo saved; QCI 8 already beats Starter by a wide margin; sufficient for most commuters
Cons: Smaller gain than QCI 9→8; add-on not universal in checkout yet; still below postpaid Turbo in our matrixCons: Upload-heavy venue routines may still choke during worst spikes

Cross-network check: if your pain is T-Mobile congestion, compare US Mobile Light Speed QCI test before you pay AT&T priority on the wrong host.


Working checklist before you add QCI 7

  1. Confirm host: Dark Star must be active—Teleport changes QoS class entirely.
  2. Own QCI 8 first: Starter (QCI 9) shoppers upgrade to Premium before considering QCI 7.
  3. Log pain on Premium: Three rush-hour upload failures at the same PIN beat one mall speed test.
  4. Check add-on availability: Line Management → Change Plan → add-ons (August 28, 2026); contact support if QCI 7 is missing.
  5. Run paired tests: Same model, same window, with and without add-on—protocol in how to check QCI level.
  6. Re-evaluate seasonally: Marcus drops QCI 7 in off-season; venue workers may keep it year-round.

Verdict

Is the US Mobile Dark Star QCI 7 upgrade worth $10/mo? Yes—if you already have Unlimited Premium (QCI 8) and can reproduce upload, jitter, or live-video failures at predictable loaded cells. No—if you are still on Starter (QCI 9), live on quiet suburban towers, or your slowness tracks weak signal rather than congestion. The QCI 8→7 step is a refinement—our August 2026 matrix shows ~1.3–1.6× upload and 40–55% loss reduction at venues, not a guaranteed win over every phone in the sector.

For Marcus, the event photographer, Premium + QCI 7 is the right stack. For Priya, Premium alone solved the garage. For Jordan in Boise, neither add-on nor Premium upgrade is rational. Your decision should follow congestion receipts on QCI 8, not forum averages or the integer alone.


Primary sources

FAQ

Short answers; details are in the article above.

How much does the US Mobile Dark Star QCI 7 upgrade cost?
As of August 28, 2026, US Mobile lists the QCI 7 priority add-on at $10 per line per month on top of Unlimited Premium (which includes QCI 8 on Dark Star). The add-on appears under Line Management → Change Plan → add-ons for eligible accounts. VIP beta customers may have had earlier access; not every checkout shows QCI 7 yet—confirm in your dashboard before you port.
Is QCI 7 faster than QCI 8 on Dark Star when towers are empty?
Usually not in a meaningful way. QCI 7 changes scheduling weight when the AT&T sector is congested. Our August 2026 off-peak controls (N=12) showed QCI 7 and QCI 8 lines overlapping within speed-test noise on quiet suburban and exurban cells. Pay for the upgrade when busy-hour upload and loss pain is reproducible—not for idle-tower peak Mbps bragging rights.
Who should buy the QCI 7 add-on instead of staying on QCI 8 Premium?
Buy QCI 7 when you already have Dark Star Unlimited Premium (QCI 8) and still lose upload throughput, jitter, or live-video stability at predictable loaded cells—stadium egress, airport concourses, hospital garages at shift change. Skip QCI 7 if you are on Starter (QCI 9)—upgrade to Premium first. Skip both if weak AT&T coverage, not congestion, is your bottleneck.
How does Dark Star QCI 7 compare to AT&T postpaid Turbo?
In our August 2026 three-way matrix, AT&T Unlimited Premium + Turbo (QCI 7 reference) edged Dark Star QCI 7 on upload at stadium egress by roughly 10–15%—both crushed QCI 8 Premium on loss and jitter. Dark Star QCI 7 is the MVNO-native path without a postpaid account; Turbo still wins when you need every last upload Mbps at venues.
How do I verify my line is on QCI 7 after adding the upgrade?
Match your plan SKU and add-on receipt to US Mobile's published QCI map, then run paired rush-hour tests at fixed locations. Android engineering menus sometimes expose QCI; iPhone does not as of August 2026. Congestion-window behavior—upload, loss, jitter—beats a single field-test screenshot.