Skip to content
type1.science

Diabeloop DBLG2

Diabeloop SA

What it is

Smartphone-based Diabeloop control algorithm, open for prescription in Germany since July 2026 paired with ViCentra's Kaleido patch pump and the Dexcom G7. The algorithm runs on the user's own Android phone (an iOS version is "coming soon"), so there is no dedicated handset to carry. It builds on the DBLG1 self-learning algorithm, which raised time-in-range ~9 points in its pivotal trial. The Netherlands follows later in summer 2026; in the US it is FDA-cleared (Dec 2025) as an interoperable automated glycemic controller but not yet in any shipping product, with a twiist integration targeting a 2027 launch.

Editorial review: .

Most recent recorded citation date: 2026-07-10. Only explicit date metadata is included; an undated citation may be newer. This does not mean every claim was reviewed on that date.

Trial status, labels and access can change between reviews. How we review the evidence · How to read the evidence

Full sources ↓Full discussion ↓Report an issue →

Germany only (prescription)Strong evidencehybrid-closed-loopphone-controltubedinteroperableself-learningsingle-hormoneOfficial site ↗

Evidence behind this assessment

Key evidence notes. Study results, product eligibility and access answer different questions.

Who was studied?
Study populations and analysis groups vary. Product age limits alone do not describe who was studied.
Benefit or performance
Time in range: Editorial proxy aligned with DBLG1 because no distinct scoring advantage is established. Predecessor evidence: median TIR 72.1% in 3,706 German DBLG1 users. FDA supporting studies used functionally identical algorithms, but this cohort was not the current DBLG2/Kaleido/G7 combination and does not establish superiority over another system.[3]
Important harms and treatment burden
Read the safety discussion and original sources. A missing summary does not establish safety.
Approval and country access
Open for prescription in Germany via distributors including Mediq and Spezimed; Netherlands to follow later in summer 2026. Not yet in any shipping US product.Approval, trial recruitment, local supply and funding are separate. Check the cited label or access source.
Follow-up and remaining uncertainty
Read the full discussion and original sources for follow-up duration and study limitations.

This record cites company or conference reports. Those sources may describe interim findings; their source type is shown in the source list.

Editorial score: calculation and evidence

A weighted editorial judgment on a 0–100 scale, not a probability of success or a measured treatment effect. Higher criterion scores mean more favorable assessments.

Score limitations. Clinical ratings use DBLG1 evidence with FDA-supported algorithm bridging, not a DBLG2-versus-Omnipod trial. Smartphone control is a documented change; higher scores do not establish clinical superiority.

Default calculation: 77 × 20 + 80 × 20 + 62 × 10 + 76 × 5 + 74 × 5 + 74 × 5 + 70 × 5 + 44 × 10 + 66 × 20 = 6990; divide by total weight 100. Unrounded weighted result: 69.9.

Time in range77

Editorial proxy aligned with DBLG1 because no distinct scoring advantage is established. Predecessor evidence: median TIR 72.1% in 3,706 German DBLG1 users. FDA supporting studies used functionally identical algorithms, but this cohort was not the current DBLG2/Kaleido/G7 combination and does not establish superiority over another system.[3]

Hypo protection80

Predecessor evidence: DBLG1 users in the German cohort achieved median time below 70 mg/dL of 0.9% and below 54 mg/dL of 0.1%. These are not direct DBLG2 results or a comparison with Omnipod populations.[3]

Automation level62

Editorial proxy aligned with DBLG1 because no distinct scoring advantage is established. FDA describes hybrid closed-loop basal adjustment and automatic correction boluses, with manual meal bolusing. Manufacturer missed-meal analyses concern DBLG1; they do not establish an automation advantage of DBLG2 over DBLG1.[7]

Average glucose76

Predecessor evidence: median GMI 7.0% in the German DBLG1 cohort. This is an editorial proxy for DBLG2, not a measured DBLG2 mean glucose or a head-to-head comparison with other products.[3]

Low variability74

Editorial proxy aligned with DBLG1 because no distinct scoring advantage is established. Predecessor evidence: mean CV 29.6% in 62 Spanish DBLG1 users. No directly comparable DBLG2-specific CV was identified; GMI is an average-glucose measure and cannot establish variability.[5]

Exercise handling74

Editorial proxy aligned with DBLG1 because no distinct scoring advantage is established. Predecessor evidence: DBLG1 post-hoc analysis found mean time below 70 mg/dL of 2.0% on activity days versus 2.2% on other days, with more preventive carbohydrate. No DBLG2-specific exercise advantage over DBLG1 or Omnipod was established.[8]

Customizability70

Editorial proxy aligned with DBLG1 because no distinct scoring advantage is established. FDA documents a 100–130 mg/dL target range, adjustable high/low thresholds and separate insulin-aggressiveness controls, plus Zen and night modes. These are documented settings, not proof of greater benefit or of added customization over DBLG1. DBLG1 already has the same core target, threshold, aggressiveness and Zen controls.[7]

Access & cost44

Open for prescription in Germany since July 2026 (Netherlands next), with the Kaleido pump and Dexcom G7 — but that is one country and one pump pairing so far, and no source establishes statutory-insurance coverage. FDA-cleared in the US but not yet in any shipping product; twiist integration targeted end-2026, US launch 2027.[1]

Freedom (form factor)66

The dedicated handset is gone — the algorithm now runs on the user's own Android phone. The Kaleido is a small body-worn "semi-patch" pump that sticks to the skin with a short tube to a nearby infusion site, so it is lighter than a belt-clipped pump but not tubeless.[2]

AID default: 50% glucose outcomes and low-glucose protection, 40% daily experience, 10% access and cost. Freedom includes tubing, wearability, water-use limits and controller requirements. These are editorial priorities; studies and device generations differ. Check each scorecard for its evidence and limits.

The full picture

Diabeloop DBLG2 is not a pump or sensor — it is the control algorithm, the "brain" of an automated insulin delivery (AID) system, delivered as a smartphone app. It reads continuous glucose monitor (CGM) values every few minutes and automatically tells a connected insulin pump how much to deliver, aiming to keep glucose in a healthy target range while reducing the daily mental load of type 1 diabetes.12

Components. DBLG2 is "interoperable": it is cleared to pair with compatible CGMs and pumps rather than locked to one device. The version now shipping in Germany pairs DBLG2 with the Dexcom G7 sensor and ViCentra's Kaleido pump — a small body-worn "semi-patch" pump that sticks to the skin with a short tube running to a nearby infusion site (like Tandem's Mobi, it is compact but not tubeless).2 The notable change from DBLG1 is that the dedicated handset is gone: the algorithm runs as an app on the user's own Android smartphone, with an iOS version described as "coming soon".2 In the US, Diabeloop has partnered with Sequel Med Tech to add DBLG2 as a second algorithm option on the twiist pump (alongside the Tidepool Loop algorithm twiist ships with today).13 twiist is a small tubed pump controlled from an iPhone and Apple Watch that uses sound waves to measure each insulin micro-dose directly.14

Clinical evidence and its limits. FDA’s DBLG2 assessment used six prospective studies with algorithms it described as functionally identical to the marketed device. That supports regulatory bridging; it is not a new comparison against Omnipod or proof that DBLG2 improves on DBLG1.7 The DBLG1 WP7 crossover trial randomized 68 adults; 63 completed both 12-week treatment periods and entered the modified intention-to-treat analysis. In that analysis, mean TIR was 68.5% with DBLG1 versus 59.4% with a sensor-augmented pump.15 The ratings here primarily use DBLG1 real-world cohorts. The July 2026 launch release’s TIR above 77% and TBR of 0.2% lack a stated cohort, size and duration, so are not rating inputs.1 A December 2025 paper cited in the launch-era discussion studied DBLG1 software versions: 1,212 users had median TIR of 68.8% before and 70.8% after the v1.16-to-v1.17 upgrade. It does not show DBLG2 adds more than three TIR points over DBLG1.9

Automation level. DBLG2 adjusts basal insulin and can deliver automatic correction boluses. FDA describes a hybrid loop with manual bolusing for meals.7 Diabeloop’s explanation advises meal announcements. Its missed-meal analyses were performed with DBLG1 in Europe; they cannot establish a unique DBLG2 feature or a fully closed-loop indication.16

Exercise. Announced activity changes insulin delivery and may prompt preventive carbohydrate. The frequently cited WP7 exercise analysis used DBLG1, not DBLG2. Similar low-glucose exposure on activity and non-activity days occurred alongside greater carbohydrate intake, so the study does not establish superiority over another system’s exercise mode.8

Ages & indications. FDA’s DBLG2 indication covers type 1 diabetes from age 12. This is the algorithm’s indication, not a statement that a particular US pump integration is on sale. Younger-child, pregnancy and type 2 indications remain development plans.717

Access. The DBLG2 algorithm was CE-marked under EU MDR in August 2025 and FDA-cleared in December 2025 as a Class II "interoperable automated glycemic controller," with a Predetermined Change Control Plan that lets it add compatible devices without a new submission.17 The integrated system — DBLG2 + Dexcom G7 + Kaleido — was CE-marked on 11 May 2026,2 and on 10 July 2026 Diabeloop and ViCentra announced it is open for prescription in Germany, sold through distributors including Mediq and Spezimed.1 The Netherlands was described as following later in summer 2026, with no posted go-live.1 Two honest caveats: this is one country and one pump pairing so far, and the announcements establish prescribability, not statutory-insurance reimbursement — check with your insurer. In the US, DBLG2 is cleared but not yet in any shipping product.

What's coming. The Netherlands launch has not been confirmed as live, and an iOS version of the app is "coming soon" (Android only at launch).2 In the US, the Sequel partnership targets product readiness by end of 2026 and a launch in 2027 — the first Diabeloop algorithm in a commercial US pump.13 Also on the roadmap: automatic exercise detection, and indication extensions to young children (4+), pregnancy, and type 2 diabetes.17

References

What's next for this

  • →Netherlands launch of DBLG2 with Kaleido and Dexcom G7 · Company guidance says later in summer 2026; a live launch was not confirmed in the sources read for this review
  • →iOS version of the app (Android only at launch) · coming soon
  • →Sequel partnership to integrate DBLG2 as a second algorithm option on the twiist pump; product readiness targeted · end of 2026
  • →First Diabeloop algorithm in a commercial US pump at US launch · 2027
  • →Automatic exercise/activity detection (still in development)
  • →Indication extensions to children 4+, pregnancy, and type 2 diabetes

Sources

  1. [1]
    Diabeloop and ViCentra Launch Smartphone-Controlled DBLG2 System Integrated with Kaleido Insulin Patch Pump and Dexcom G7 CGM, Now Open for Prescription · Manufacturer · 2026-07-10 — Launch announcement. Its unqualified "TIR >77% / TBR 0.2%" real-world figure has no stated cohort, size or duration, so it is not used to score this system.

    Diabeloop. Diabeloop and ViCentra Launch Smartphone-Controlled DBLG2 System Integrated with Kaleido Insulin Patch Pump and Dexcom G7 CGM, Now Open for Prescription. Diabeloop Press Release (10 July 2026).

  2. [2]
    Diabeloop Announces CE Marking and Gradual Launch of the New DBLG2 System, Integrated with Dexcom G7 CGM and ViCentra's Kaleido Insulin Patch Pump · Manufacturer · 2026-05-11

    Diabeloop. Diabeloop Announces CE Marking and Gradual Launch of the New DBLG2 System, Integrated with Dexcom G7 CGM and ViCentra's Kaleido Insulin Patch Pump. Diabeloop Press Release (11 May 2026).

  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
    FDA DBLG2 K251152 decision summary: hybrid-loop indication, settings and clinical bridging · Regulatory decision

    US FDA. DBLG2 K251152 decision summary, including indication, configurable settings and clinical bridging.

  8. [8]
    DBLG1 WP7 physical-activity post-hoc analysis · Peer-reviewed study

    Franc S, Benhamou PY, Borot S, et al. No more hypoglycaemia on days with physical activity and unrestricted diet when using a closed-loop system for 12 weeks: A post hoc secondary analysis of the multicentre, randomized controlled Diabeloop WP7 trial. Diabetes Obes Metab (2021).

  9. [9]
    DBLG1 algorithm versions: two-step observational report · Peer-reviewed study

    Benhamou et al. DBLG1 algorithm modifications: a two-step observational report (18 December 2025).

  10. [10]
  11. [11]
  12. [12]

    Diabeloop. DBLG2: an automated insulin delivery algorithm in an app — FDA 510(k) clearance. Diabeloop News (2026). https://www.diabeloop.com/news/company/diabeloop-fda-clearance-en

  13. [13]

    Diabeloop. Diabeloop and Sequel Med Tech enter partnership to integrate FDA-cleared DBLG2 into the twiist Automated Insulin Delivery System. Diabeloop Press Release (2026). https://www.diabeloop.com/media-press/press-releases/dbl-sequel-pr-en

  14. [14]

    Sequel Med Tech. Sequel and Diabeloop enter partnership to integrate DBLG2 into twiist. Sequel Med Tech News (2026). https://www.sequelmedtech.com/blog-posts/sequel-and-diabeloop-enter-partnership-to-integrate-into-twiist

  15. [15]

    Benhamou PY, Franc S, Reznik Y, et al. Closed-loop insulin delivery in adults with type 1 diabetes in real-life conditions: a 12-week multicentre, open-label randomised controlled crossover trial. Lancet Digit Health (2019). https://doi.org/10.1016/S2589-7500%2819%2930003-2

  16. [16]

    Diabeloop. DBLG2: FDA-cleared and CE-marked advanced AID algorithm designed to manage unexpected glycemic excursions. Diabeloop News (2026). https://www.diabeloop.com/news/health-sector/dblg2-manage-unexpected-glycemic-excursions

  17. [17]

    Diabeloop. Diabeloop receives FDA 510(k) clearance for DBLG2: a strategic turning point. Diabeloop Press Release (2026). https://www.diabeloop.com/media-press/press-releases/diabeloop-fda-510k-clearance