diff --git a/PWGLF/Tasks/Nuspex/CMakeLists.txt b/PWGLF/Tasks/Nuspex/CMakeLists.txt
index a1db5c5f0f4..dff5c8a048b 100644
--- a/PWGLF/Tasks/Nuspex/CMakeLists.txt
+++ b/PWGLF/Tasks/Nuspex/CMakeLists.txt
@@ -201,6 +201,11 @@ o2physics_add_dpl_workflow(multiplicity-pt
     PUBLIC_LINK_LIBRARIES O2Physics::AnalysisCore
     COMPONENT_NAME Analysis)
 
+o2physics_add_dpl_workflow(flattenicity-pt-spectra
+                    SOURCES flattenicityPtSpectra.cxx
+                    PUBLIC_LINK_LIBRARIES O2::Framework O2Physics::AnalysisCore
+                    COMPONENT_NAME Analysis)
+
 o2physics_add_dpl_workflow(deuteron-in-jets-trg-pt
     SOURCES DeuteronInJetsTrgPt.cxx
     PUBLIC_LINK_LIBRARIES O2Physics::AnalysisCore O2Physics::PWGJECore FastJet::FastJet FastJet::Contrib O2Physics::EventFilteringUtils
diff --git a/PWGLF/Tasks/Nuspex/flattenicityPtSpectra.cxx b/PWGLF/Tasks/Nuspex/flattenicityPtSpectra.cxx
new file mode 100644
index 00000000000..aa85b3ed05a
--- /dev/null
+++ b/PWGLF/Tasks/Nuspex/flattenicityPtSpectra.cxx
@@ -0,0 +1,311 @@
+// Copyright 2019-2020 CERN and copyright holders of ALICE O2.
+// See https://alice-o2.web.cern.ch/copyright for details of the copyright holders.
+// All rights not expressly granted are reserved.
+//
+// This software is distributed under the terms of the GNU General Public
+// License v3 (GPL Version 3), copied verbatim in the file "COPYING".
+//
+// In applying this license CERN does not waive the privileges and immunities
+// granted to it by virtue of its status as an Intergovernmental Organization
+// or submit itself to any jurisdiction.
+///
+/// \author Dushmanta Sahu  (dushmanta.sahus@cern.ch)
+/// \since September 1, 2026
+/// \file flattenicityPtSpectra.cxx
+/// \brief  Analysis to do flattenicity pt spectra
+
+#include "Common/DataModel/EventSelection.h"
+#include "Common/DataModel/FT0Corrected.h"
+#include "Common/DataModel/Multiplicity.h"
+#include "Common/DataModel/TrackSelectionTables.h"
+
+#include <CommonConstants/MathConstants.h>
+#include <Framework/AnalysisDataModel.h>
+#include <Framework/AnalysisTask.h>
+#include <Framework/runDataProcessing.h>
+
+#include <TH1F.h>
+#include <TH2F.h>
+#include <TH3F.h>
+#include <TMath.h>
+
+#include <algorithm>
+#include <array>
+#include <cmath>
+#include <vector>
+
+using namespace o2;
+using namespace o2::framework;
+
+using MyTracks = soa::Join<aod::Tracks, aod::TracksExtra, aod::TracksDCA>;
+
+using MyCollisions = soa::Join<aod::Collisions,
+                               aod::Mults,
+                               aod::FT0sCorrected,
+                               aod::EvSels>;
+
+struct FlattenicityPtSpectra {
+
+  Configurable<float> etaMin{"etaMin", -0.8f, "Min eta for mid-rapidity tracks"};
+  Configurable<float> etaMax{"etaMax", 0.8f, "Max eta for mid-rapidity tracks"};
+  Configurable<float> ptMin{"ptMin", 0.15f, "Min pT (GeV/c)"};
+  Configurable<float> ptMax{"ptMax", 20.0f, "Max pT (GeV/c)"};
+  Configurable<int> minTPCnClsFound{"minTPCnClsFound", 70, "Min TPC found clusters"};
+  Configurable<float> maxDCAz{"maxDCAz", 2.0f, "Max |DCAz| (cm)"};
+  Configurable<float> maxVtxZ{"maxVtxZ", 10.0f, "Max |vtx z| (cm)"};
+
+  Configurable<int> minActiveFT0Ch{"minActiveFT0Ch", 4, "Min active FT0 channels for flattenicity"};
+
+  static constexpr int NChA = 96;
+  static constexpr int NChC = 96;
+  static constexpr int NCells = NChA + NChC; // 192 total
+
+  HistogramRegistry registry{
+    "registry",
+    {},
+    OutputObjHandlingPolicy::AnalysisObject,
+    true,
+    true};
+
+  template <typename TrackType>
+  bool isGoodTrack(TrackType const& track) const
+  {
+    if (!track.hasTPC())
+      return false;
+    if (track.tpcNClsFound() < minTPCnClsFound.value)
+      return false;
+    if (track.eta() < etaMin.value || track.eta() > etaMax.value)
+      return false;
+    if (std::abs(track.dcaZ()) > maxDCAz.value)
+      return false;
+    if (track.pt() < ptMin.value || track.pt() > ptMax.value)
+      return false;
+    return true;
+  }
+
+  float calculateFlattenicityFT0(aod::FT0 const& ft0, int minActive) const
+  {
+    std::array<float, NCells> signals{};
+    signals.fill(0.f);
+
+    int nActive = 0;
+
+    int chIdx = 0;
+    for (const auto& a : ft0.amplitudeA()) {
+      if (chIdx < NChA) {
+        float amp = static_cast<float>(a);
+        if (amp > 0.f) {
+          signals[chIdx] = amp;
+          ++nActive;
+        }
+      }
+      ++chIdx;
+    }
+
+    chIdx = 0;
+    for (const auto& a : ft0.amplitudeC()) {
+      if (chIdx < NChC) {
+        float amp = static_cast<float>(a);
+        if (amp > 0.f) {
+          signals[NChA + chIdx] = amp;
+          ++nActive;
+        }
+      }
+      ++chIdx;
+    }
+
+    if (nActive < minActive)
+      return -1.f;
+
+    float mRho = 0.f;
+    for (int i = 0; i < NCells; ++i) {
+      mRho += signals[i];
+    }
+    mRho /= static_cast<float>(NCells);
+    if (mRho <= 0.f)
+      return -1.f;
+
+    float sRhoTmp = 0.f;
+    for (int i = 0; i < NCells; ++i) {
+      if (signals[i] > 0.f) {
+        float d = signals[i] - mRho;
+        sRhoTmp += d * d;
+      }
+    }
+    sRhoTmp /= static_cast<float>(NCells) * static_cast<float>(NCells);
+    float sRho = std::sqrt(sRhoTmp);
+
+    // flattenicity = 1 - sigma/mean, clamped to [0, 1]
+    float cv = sRho / mRho;
+    return std::max(0.f, std::min(1.f, 1.f - cv));
+  }
+
+  void init(InitContext const&)
+  {
+    AxisSpec ptAxis{200, 0.0f, 20.0f, "#it{p}_{T} (GeV/c)"};
+    AxisSpec flatAxis{102, -0.01f, 1.01f, "Flattenicity"};
+    AxisSpec multAxis{300, 0.f, 3000.f, "FT0M amplitude (a.u.)"};
+    AxisSpec nchAxis{500, -0.5f, 499.5f, "Raw N_{ch} (|#eta|<0.8, p_{T}>0.15)"};
+    AxisSpec vtxzAxis{200, -20.f, 20.f, "v_{z} (cm)"};
+
+    registry.add("hEventCounter",
+                 "Event counter",
+                 HistType::kTH1F, {{5, 0.5, 5.5}});
+    registry.get<TH1>(HIST("hEventCounter"))->GetXaxis()->SetBinLabel(1, "All");
+    registry.get<TH1>(HIST("hEventCounter"))->GetXaxis()->SetBinLabel(2, "vtxZ");
+    registry.get<TH1>(HIST("hEventCounter"))->GetXaxis()->SetBinLabel(3, "hasFT0");
+    registry.get<TH1>(HIST("hEventCounter"))->GetXaxis()->SetBinLabel(4, "FT0signal");
+    registry.get<TH1>(HIST("hEventCounter"))->GetXaxis()->SetBinLabel(5, "goodFlat");
+
+    registry.add("hVtxZ",
+                 "Primary vertex z",
+                 HistType::kTH1F, {vtxzAxis});
+
+    registry.add("hFT0MMultiplicity",
+                 "FT0M amplitude sum",
+                 HistType::kTH1F, {multAxis});
+
+    registry.add("hFlattenicityDistribution",
+                 "Flattenicity (Gyula method, 192 FT0 ch)",
+                 HistType::kTH1F, {flatAxis});
+
+    registry.add("hFT0AAmplitude",
+                 "Total FT0A amplitude",
+                 HistType::kTH1F, {{200, 0.f, 2000.f, "Amplitude (a.u.)"}});
+
+    registry.add("hFT0CAmplitude",
+                 "Total FT0C amplitude",
+                 HistType::kTH1F, {{200, 0.f, 2000.f, "Amplitude (a.u.)"}});
+
+    // QA: active channel count — use this to tune minActiveFT0Ch
+    registry.add("hNActiveFT0Channels",
+                 "Active FT0 channels per event (amp > 0)",
+                 HistType::kTH1F, {{193, -0.5f, 192.5f, "N active channels"}});
+
+    registry.add("hEtaDistribution",
+                 "Eta distribution (selected tracks)",
+                 HistType::kTH1F, {{100, -1.0f, 1.0f, "#eta"}});
+
+    registry.add("hPhiDistribution",
+                 "Phi distribution (selected tracks)",
+                 HistType::kTH1F, {{100, 0.f, o2::constants::math::TwoPI, "#phi"}});
+
+    registry.add("hPtSpectrum_All",
+                 "pT spectrum (all selected events)",
+                 HistType::kTH1F, {ptAxis});
+
+    registry.add("hFT0MMultVsFlattenicity",
+                 "FT0M multiplicity vs Flattenicity;"
+                 "FT0M amplitude (a.u.);Flattenicity",
+                 HistType::kTH2F, {multAxis, flatAxis});
+
+    registry.add("hFT0MMultVsFlattenicityVsPt",
+                 "FT0M multiplicity vs Flattenicity vs #it{p}_{T};"
+                 "FT0M amplitude (a.u.);Flattenicity;#it{p}_{T} (GeV/c)",
+                 HistType::kTH3F, {multAxis, flatAxis, ptAxis});
+
+    registry.add("hNchDistribution",
+                 "Raw charged particle multiplicity",
+                 HistType::kTH1F, {nchAxis});
+
+    registry.add("hNchVsFlattenicity",
+                 "Raw N_{ch} vs Flattenicity;"
+                 "Raw N_{ch};Flattenicity",
+                 HistType::kTH2F, {nchAxis, flatAxis});
+
+    registry.add("hNchVsFlattenicityVsPt",
+                 "Raw N_{ch} vs Flattenicity vs #it{p}_{T};"
+                 "Raw N_{ch};Flattenicity;#it{p}_{T} (GeV/c)",
+                 HistType::kTH3F, {nchAxis, flatAxis, ptAxis});
+
+    LOG(info) << "FlattenicityPtSpectra::init — done"
+              << "  NCells=" << NCells
+              << "  minActiveFT0Ch=" << minActiveFT0Ch.value
+              << "  Flattenicity method: Gyula (fixed denominator = total channels)";
+  }
+
+  void process(MyCollisions::iterator const& collision,
+               MyTracks const& tracks,
+               aod::FT0s const& /*ft0s*/)
+  {
+    registry.fill(HIST("hEventCounter"), 1.f);
+
+    if (std::abs(collision.posZ()) > maxVtxZ.value)
+      return;
+    registry.fill(HIST("hEventCounter"), 2.f);
+    registry.fill(HIST("hVtxZ"), collision.posZ());
+
+    if (!collision.has_foundFT0())
+      return;
+    registry.fill(HIST("hEventCounter"), 3.f);
+
+    auto ft0 = collision.foundFT0();
+
+    float sumA = 0.f, sumC = 0.f;
+    int nActive = 0;
+    for (const auto& a : ft0.amplitudeA()) {
+      if (a > 0.f) {
+        sumA += a;
+        ++nActive;
+      }
+    }
+    for (const auto& a : ft0.amplitudeC()) {
+      if (a > 0.f) {
+        sumC += a;
+        ++nActive;
+      }
+    }
+
+    registry.fill(HIST("hFT0AAmplitude"), sumA);
+    registry.fill(HIST("hFT0CAmplitude"), sumC);
+    registry.fill(HIST("hNActiveFT0Channels"), static_cast<float>(nActive));
+
+    if (sumA + sumC <= 0.f)
+      return;
+    registry.fill(HIST("hEventCounter"), 4.f);
+
+    float flattenicity = calculateFlattenicityFT0(ft0, minActiveFT0Ch.value);
+    if (flattenicity < 0.f)
+      return;
+    registry.fill(HIST("hEventCounter"), 5.f);
+
+    float multFT0M = collision.multFT0M();
+
+    registry.fill(HIST("hFT0MMultiplicity"), multFT0M);
+    registry.fill(HIST("hFlattenicityDistribution"), flattenicity);
+    registry.fill(HIST("hFT0MMultVsFlattenicity"), multFT0M, flattenicity);
+
+    std::vector<typename MyTracks::iterator> goodTracks;
+    goodTracks.reserve(1000);
+
+    for (const auto& track : tracks) {
+      if (!isGoodTrack(track))
+        continue;
+      goodTracks.push_back(track);
+    }
+
+    int rawNch = goodTracks.size();
+
+    // Fill per-event histograms with Nch
+    registry.fill(HIST("hNchDistribution"), rawNch);
+    registry.fill(HIST("hNchVsFlattenicity"), rawNch, flattenicity);
+
+    // Fill per-track histograms using the collected good tracks
+    for (const auto& track : goodTracks) {
+      registry.fill(HIST("hEtaDistribution"), track.eta());
+      registry.fill(HIST("hPhiDistribution"), track.phi());
+      registry.fill(HIST("hPtSpectrum_All"), track.pt());
+      registry.fill(HIST("hFT0MMultVsFlattenicityVsPt"),
+                    multFT0M, flattenicity, track.pt());
+
+      // NEW: Fill Nch-based 3D histogram
+      registry.fill(HIST("hNchVsFlattenicityVsPt"),
+                    rawNch, flattenicity, track.pt());
+    }
+  }
+};
+
+WorkflowSpec defineDataProcessing(ConfigContext const& cfgc)
+{
+  return WorkflowSpec{adaptAnalysisTask<FlattenicityPtSpectra>(cfgc)};
+}