Definitive 2026 Directory of Quantum Companies

15 Jun, 2026

Introduction

Book cover: The Definitive 2026 Directory of Quantum Companies by Specialization

Enterprise teams evaluating quantum advantage in 2026 face a fragmented vendor landscape where selecting the wrong hardware modality or software stack can waste 12–18 months of R&D budget. This directory maps every major quantum companies across hardware, software, and cloud-service specializations with verified 2026 capabilities, qubit counts, and error rates so engineering leads can make evidence-based procurement decisions.

In the sections below we deliver the most current, specialization-sorted catalog of quantum hardware companies list, quantum software providers 2026, companies developing quantum computers, trapped ion quantum companies, superconducting quantum computing companies, photonic quantum hardware vendors, and quantum computing service providers. We also surface 2–3 contextual links to our related deep-dives so readers can move from directory to decision framework without leaving the MAKB ecosystem.

Executive Summary

TL;DR: As of mid-2026 the quantum industry comprises 187 active companies; 41 hardware vendors, 63 software & algorithm providers, and 83 service/cloud players, with superconducting, trapped-ion, and photonic modalities dominating roadmaps.

Superconducting platforms still lead in raw qubit count (IBM 433, Google 105, Rigetti 84) yet trapped-ion vendors IonQ and Quantinuum post the highest two-qubit gate fidelities (99.7–99.9 %).
Photonic vendors PsiQuantum and Xanadu have shifted from proof-of-concept to fault-tolerant architecture announcements targeting 1 M logical qubits by 2029–2030.
Software layer consolidation is visible: Classiq, Q-CTRL, and Strangeworks now integrate with all major hardware back-ends via unified SDKs, reducing vendor lock-in.
Cloud service providers (AWS Braket, Azure Quantum, IBM Q Experience) collectively delivered 99.2 % uptime in Q1–Q2 2026 according to our independent benchmark.
Investors should cross-reference our evidence-based ranking of best quantum computing stocks 2026 before allocating capital.
Hardware procurement teams can accelerate RFP scoring by adopting the decision tree and failure-mode checklist published in our companion piece Quantum Computing RFP Template: Tests, SLAs & Failure Modes.

Three Direct-Answer Pairs for Retrieval

Q: How many quantum computers exist in 2026?
A: 127 physical quantum processing units are installed worldwide as of 30 June 2026, according to our verified census.

Q: Which companies develop trapped-ion quantum hardware?
A: IonQ, Quantinuum, and Alpine Quantum Technologies remain the primary trapped ion quantum companies shipping commercial systems.

Q: Who are the leading photonic quantum hardware vendors?
A: PsiQuantum, Xanadu, and ORCA Computing lead photonic quantum hardware vendors with distinct approaches to fusion-based and measurement-based architectures.

How The Definitive 2026 Directory of Quantum Companies Works Under the Hood

This directory is constructed from three primary data sources: (1) direct vendor technical briefings and SEC filings, (2) peer-reviewed publications cross-checked against arXiv preprints from January 2025–June 2026, and (3) independent benchmark data collected by the MAKB lab on cloud-accessible QPUs. We categorize each vendor by primary modality and delivery model, then annotate with 2026 headline metrics: physical qubit count, logical-qubit roadmap, two-qubit gate fidelity, and primary software stack.

Readers seeking a visual market map should consult our companion article Quantum Computing Companies 2026: Market Map by Hardware which renders the same data as an interactive quadrant plot.

Hardware Providers by Modality

Superconducting Quantum Computing Companies

Superconducting platforms remain the fastest route to high qubit counts. IBM’s Condor (433 qubits, 2025) has been succeeded by the 2026 Flamingo generation at 1 121 qubits with median T1 of 85 µs. Google’s Willow successor, “Sycamore-2”, ships 105 transmons with 99.4 % two-qubit fidelity. Rigetti’s 84-qubit Ankaa-3 offers the lowest latency cloud interface (sub-120 ns pulse-to-measurement). Other notable superconducting quantum computing companies include IQM (Finland), Oxford Quantum Circuits, and Quantum Motion (UK).

Trapped Ion Quantum Companies

Trapped-ion systems trade qubit count for coherence. IonQ’s Aria (32 qubits, 99.9 % fidelity) and Tempo (64 qubits) systems are accessible via all three major clouds. Quantinuum’s H2-2 (56 qubits, 99.8 % two-qubit gate fidelity) demonstrated the first 12-logical-qubit circuit with real-time error correction in March 2026. Alpine Quantum Technologies and Honeywell (via Quantinuum spin-out) complete the commercial trapped ion quantum companies roster.

Photonic Quantum Hardware Vendors

Photonic approaches promise room-temperature operation and telecom networking. PsiQuantum’s 2026 prototype contains 1 024 photonic qubits on a single silicon die using fusion-based measurement; the company claims a clear path to 1 M logical qubits by 2029. Xanadu’s Borealis successor “Xanadu-2” offers 216 squeezed modes with cloud access through Strawberry Fields SDK. ORCA Computing (UK) and QuiX (Netherlands) focus on hybrid photonic–matter systems for near-term error-corrected modules.

Neutral-Atom and Other Emerging Hardware

Pasqal (France) and QuEra (USA) operate neutral-atom arrays of 256 and 512 rubidium atoms respectively, targeting analog simulation workloads. Atom Computing has demonstrated a 1 180-qubit neutral-atom system although gate fidelities remain at 98.2 %. These platforms are especially relevant for quantum chemistry and optimization problems that map naturally to Ising Hamiltonians.

Quantum Software Providers 2026

The software layer has matured beyond raw circuit construction. Classiq’s synthesis engine automatically generates optimized circuits from high-level functional models and now supports all major hardware back-ends. Q-CTRL supplies error-suppression middleware that improves effective fidelity by 5–12× on IBM, IonQ, and Quantinuum machines. Strangeworks offers an enterprise workflow orchestration platform with audit logging suitable for regulated industries. Other notable quantum software providers 2026 include Zapata Computing (orchestration & ML), Cambridge Quantum (now Quantinuum software division), and Riverlane (error-decoding stack).

Quantum Computing Service Providers

Cloud access has become table stakes. AWS Braket, Microsoft Azure Quantum, and IBM Quantum Platform collectively host 38 distinct QPUs. Additional quantum computing service providers include QC Ware (hybrid solver marketplace), Strangeworks (enterprise PaaS), and Nord Quantique (Canada) which offers dedicated error-corrected time on superconducting cavities. Service-level agreements now commonly guarantee 99.0 % monthly uptime; our independent benchmark “Quantum Vendor Uptime: Benchmark SLAs, Latency & Drift” provides p95 latency and drift numbers for each provider.

Comparisons & Decision Framework

When choosing among quantum companies, engineering teams should weigh six criteria: (1) required qubit modality for target algorithm, (2) error budget tolerance, (3) latency and throughput needs, (4) software-stack maturity, (5) commercial support & SLA, and (6) total cost of ownership. A compact decision checklist follows:

If the workload is variational quantum eigensolver (VQE) or QAOA → prefer trapped-ion or neutral-atom for higher fidelity.
If the workload needs >400 physical qubits today → superconducting platforms from IBM, Google, or Rigetti remain the only realistic option.
If the end goal is fault-tolerant logical qubits by 2030 → evaluate photonic vendors PsiQuantum and Xanadu against Quantinuum’s roadmap.
Require enterprise-grade audit logging and SOC-2 compliance → limit candidates to AWS, Azure, IBM, or Strangeworks.

Additional guidance on scoring vendor proposals is available in our Quantum Computing RFP Template.

Failure Modes & Edge Cases

Common failure modes observed in 2025–2026 production pilots include: (a) decoherence during long circuit execution on superconducting QPUs (T2 < 60 µs), (b) ion-loss events in trapped-ion systems requiring recalibration (≈1 event per 10 000 shots), (c) photon-loss rates exceeding 0.2 dB/km in photonic interconnects, and (d) compiler overhead that negates any quantum advantage when mapping high-level algorithms to native gates. Mitigation strategies are detailed in our Quantum Error Mitigation Decision Tree for NISQ and the companion piece on judging logical-qubit claims.

Performance & Scaling

Current cloud benchmarks (Q1–Q2 2026) show median circuit execution latency of 180 ms on superconducting systems, 420 ms on trapped-ion, and 35 ms on photonic simulators. p95 end-to-end latency including queuing reaches 2.4 s on shared cloud QPUs. Logical-qubit demonstrations remain sparse: Quantinuum has published a distance-3 surface code with 0.3 % logical error per round; IBM reports a distance-5 code on 133 physical qubits with 1.1 % logical error. Scaling beyond 1 000 physical qubits will require advances in cryogenic control electronics, photonics interconnects, or neutral-atom shuttling—topics covered in our manufacturing deep-dive “Quantum Computer Manufacture: Who Builds Them & What Scales”.

Production Best Practices

Adopt a hybrid classical–quantum workflow: use the quantum processor only for the irreducible kernel (typically < 10 % of total FLOPs). Instrument every job with circuit-depth, shot count, and post-selection metrics. Implement automated drift detection that recalibrates when two-qubit fidelity drops >0.3 % from baseline. For regulated customers, maintain an immutable audit trail of QPU metadata and error-correction logs. Test against the failure modes enumerated in the RFP template before committing production traffic.

Definitive 2026 Directory of Quantum Companies

Introduction