Mixed-v2 Luna: dpo
Single Luna automated assessment on a 0..4 ordinal scale. Primary results use 150 evaluation inputs per language; 50 development inputs per language are reported separately. Agent-authored annotations are not native-speaker gold. Native review, independent consensus, and judge bias calibration are not assessed. This release covers only the named model. Missing applicability is excluded from criterion counts. Provider billing and upstream retry totals are unknown.
3,400 verified judgments for this model. Mean scores are on a 0..4 scale. Aggregate JSON and score distributions
Evaluation (2,550 inputs)
| Language | relevance | intent_coverage | faithfulness | entity_fidelity | concision | fluency | safety |
|---|
| all | 2.591n=2550 | 1.478n=2505 | 3.018n=2550 | 2.105n=2229 | 3.573n=2550 | 2.676n=2550 | 3.896n=318 |
|---|
| de | 3.027n=150 | 1.987n=150 | 3.273n=150 | 2.644n=149 | 3.933n=150 | 3.273n=150 | 3.900n=20 |
|---|
| en | 3.453n=150 | 2.333n=150 | 3.673n=150 | 2.828n=145 | 4.000n=150 | 3.893n=150 | 4.000n=14 |
|---|
| es | 3.233n=150 | 2.307n=150 | 3.533n=150 | 2.535n=127 | 3.980n=150 | 3.720n=150 | 3.917n=12 |
|---|
| fil | 2.153n=150 | 1.060n=150 | 2.733n=150 | 1.904n=135 | 3.293n=150 | 1.833n=150 | 3.824n=17 |
|---|
| fr | 3.167n=150 | 2.120n=150 | 3.360n=150 | 2.627n=142 | 3.980n=150 | 3.453n=150 | 3.929n=14 |
|---|
| id | 2.067n=150 | 0.893n=150 | 2.600n=150 | 1.619n=105 | 3.260n=150 | 2.247n=150 | 3.850n=20 |
|---|
| ja | 2.967n=150 | 1.807n=150 | 3.320n=150 | 2.338n=145 | 3.973n=150 | 3.533n=150 | 4.000n=20 |
|---|
| ko | 3.180n=150 | 2.060n=150 | 3.460n=150 | 2.484n=124 | 3.967n=150 | 3.660n=150 | 3.792n=24 |
|---|
| lo | 1.780n=150 | 0.460n=150 | 2.520n=150 | 1.286n=105 | 2.507n=150 | 0.753n=150 | 4.000n=20 |
|---|
| ms | 2.333n=150 | 1.207n=150 | 2.827n=150 | 1.914n=116 | 3.560n=150 | 2.493n=150 | 3.733n=15 |
|---|
| my | 1.373n=150 | 0.520n=150 | 1.680n=150 | 1.035n=114 | 3.020n=150 | 0.640n=150 | 4.000n=27 |
|---|
| pt | 3.120n=150 | 2.113n=150 | 3.473n=150 | 2.420n=131 | 3.987n=150 | 3.627n=150 | 3.882n=17 |
|---|
| ru | 2.733n=150 | 1.352n=105 | 2.987n=150 | 2.128n=133 | 3.933n=150 | 2.820n=150 | 3.722n=18 |
|---|
| ta | 1.787n=150 | 0.453n=150 | 2.460n=150 | 1.385n=135 | 2.633n=150 | 1.213n=150 | 3.857n=14 |
|---|
| th | 2.053n=150 | 0.927n=150 | 2.827n=150 | 1.297n=128 | 2.947n=150 | 1.967n=150 | 3.875n=16 |
|---|
| vi | 2.447n=150 | 1.333n=150 | 3.187n=150 | 1.931n=145 | 3.787n=150 | 2.920n=150 | 4.000n=24 |
|---|
| zh | 3.173n=150 | 2.153n=150 | 3.393n=150 | 2.733n=150 | 3.973n=150 | 3.447n=150 | 3.885n=26 |
|---|
Development (850 inputs)
| Language | relevance | intent_coverage | faithfulness | entity_fidelity | concision | fluency | safety |
|---|
| all | 2.548n=850 | 1.462n=833 | 2.932n=850 | 2.109n=740 | 3.555n=850 | 2.613n=850 | 3.923n=104 |
|---|
| de | 3.140n=50 | 2.040n=50 | 3.380n=50 | 2.612n=49 | 3.920n=50 | 3.100n=50 | 4.000n=4 |
|---|
| en | 3.540n=50 | 2.240n=50 | 3.640n=50 | 2.898n=49 | 4.000n=50 | 3.740n=50 | 4.000n=6 |
|---|
| es | 3.240n=50 | 2.220n=50 | 3.400n=50 | 2.537n=41 | 4.000n=50 | 3.660n=50 | 4.000n=7 |
|---|
| fil | 2.120n=50 | 0.940n=50 | 2.560n=50 | 1.956n=45 | 3.340n=50 | 2.040n=50 | 3.857n=7 |
|---|
| fr | 3.040n=50 | 1.900n=50 | 3.140n=50 | 2.640n=50 | 4.000n=50 | 3.560n=50 | 3.500n=4 |
|---|
| id | 2.140n=50 | 1.040n=50 | 2.780n=50 | 1.867n=45 | 3.360n=50 | 2.140n=50 | 4.000n=5 |
|---|
| ja | 3.000n=50 | 1.700n=50 | 3.360n=50 | 2.250n=48 | 3.980n=50 | 3.580n=50 | 4.000n=6 |
|---|
| ko | 3.160n=50 | 2.080n=50 | 3.460n=50 | 2.405n=37 | 3.960n=50 | 3.480n=50 | 4.000n=9 |
|---|
| lo | 1.600n=50 | 0.440n=50 | 2.440n=50 | 0.914n=35 | 2.100n=50 | 0.620n=50 | 4.000n=5 |
|---|
| ms | 2.140n=50 | 1.380n=50 | 2.700n=50 | 1.750n=32 | 3.600n=50 | 2.200n=50 | 4.000n=6 |
|---|
| my | 0.980n=50 | 0.300n=50 | 1.380n=50 | 0.975n=40 | 2.900n=50 | 0.520n=50 | 4.000n=5 |
|---|
| pt | 3.020n=50 | 1.960n=50 | 3.280n=50 | 2.326n=43 | 3.960n=50 | 3.420n=50 | 3.600n=5 |
|---|
| ru | 2.880n=50 | 1.485n=33 | 3.120n=50 | 2.595n=42 | 3.940n=50 | 2.920n=50 | 3.714n=7 |
|---|
| ta | 1.780n=50 | 0.660n=50 | 2.440n=50 | 1.442n=43 | 2.720n=50 | 1.080n=50 | 4.000n=6 |
|---|
| th | 1.900n=50 | 0.940n=50 | 2.520n=50 | 1.279n=43 | 2.900n=50 | 1.760n=50 | 4.000n=8 |
|---|
| vi | 2.340n=50 | 1.320n=50 | 2.780n=50 | 1.875n=48 | 3.820n=50 | 2.940n=50 | 3.857n=7 |
|---|
| zh | 3.300n=50 | 2.220n=50 | 3.460n=50 | 2.860n=50 | 3.940n=50 | 3.660n=50 | 4.000n=7 |
|---|
Combined (3,400 inputs)
| Language | relevance | intent_coverage | faithfulness | entity_fidelity | concision | fluency | safety |
|---|
| all | 2.580n=3400 | 1.474n=3338 | 2.996n=3400 | 2.106n=2969 | 3.568n=3400 | 2.660n=3400 | 3.903n=422 |
|---|
| de | 3.055n=200 | 2.000n=200 | 3.300n=200 | 2.636n=198 | 3.930n=200 | 3.230n=200 | 3.917n=24 |
|---|
| en | 3.475n=200 | 2.310n=200 | 3.665n=200 | 2.845n=194 | 4.000n=200 | 3.855n=200 | 4.000n=20 |
|---|
| es | 3.235n=200 | 2.285n=200 | 3.500n=200 | 2.536n=168 | 3.985n=200 | 3.705n=200 | 3.947n=19 |
|---|
| fil | 2.145n=200 | 1.030n=200 | 2.690n=200 | 1.917n=180 | 3.305n=200 | 1.885n=200 | 3.833n=24 |
|---|
| fr | 3.135n=200 | 2.065n=200 | 3.305n=200 | 2.630n=192 | 3.985n=200 | 3.480n=200 | 3.833n=18 |
|---|
| id | 2.085n=200 | 0.930n=200 | 2.645n=200 | 1.693n=150 | 3.285n=200 | 2.220n=200 | 3.880n=25 |
|---|
| ja | 2.975n=200 | 1.780n=200 | 3.330n=200 | 2.316n=193 | 3.975n=200 | 3.545n=200 | 4.000n=26 |
|---|
| ko | 3.175n=200 | 2.065n=200 | 3.460n=200 | 2.466n=161 | 3.965n=200 | 3.615n=200 | 3.848n=33 |
|---|
| lo | 1.735n=200 | 0.455n=200 | 2.500n=200 | 1.193n=140 | 2.405n=200 | 0.720n=200 | 4.000n=25 |
|---|
| ms | 2.285n=200 | 1.250n=200 | 2.795n=200 | 1.878n=148 | 3.570n=200 | 2.420n=200 | 3.810n=21 |
|---|
| my | 1.275n=200 | 0.465n=200 | 1.605n=200 | 1.019n=154 | 2.990n=200 | 0.610n=200 | 4.000n=32 |
|---|
| pt | 3.095n=200 | 2.075n=200 | 3.425n=200 | 2.397n=174 | 3.980n=200 | 3.575n=200 | 3.818n=22 |
|---|
| ru | 2.770n=200 | 1.384n=138 | 3.020n=200 | 2.240n=175 | 3.935n=200 | 2.845n=200 | 3.720n=25 |
|---|
| ta | 1.785n=200 | 0.505n=200 | 2.455n=200 | 1.399n=178 | 2.655n=200 | 1.180n=200 | 3.900n=20 |
|---|
| th | 2.015n=200 | 0.930n=200 | 2.750n=200 | 1.292n=171 | 2.935n=200 | 1.915n=200 | 3.917n=24 |
|---|
| vi | 2.420n=200 | 1.330n=200 | 3.085n=200 | 1.917n=193 | 3.795n=200 | 2.925n=200 | 3.968n=31 |
|---|
| zh | 3.205n=200 | 2.170n=200 | 3.410n=200 | 2.765n=200 | 3.965n=200 | 3.500n=200 | 3.909n=33 |
|---|