diff --git a/plots/line-loss-training/implementations/altair.py b/plots/line-loss-training/implementations/altair.py
new file mode 100644
index 0000000000..9757106387
--- /dev/null
+++ b/plots/line-loss-training/implementations/altair.py
@@ -0,0 +1,102 @@
+""" pyplots.ai
+line-loss-training: Training Loss Curve
+Library: altair 6.0.0 | Python 3.13.11
+Quality: 91/100 | Created: 2025-12-31
+"""
+
+import altair as alt
+import numpy as np
+import pandas as pd
+
+
+# Data - Simulating neural network training loss curves
+np.random.seed(42)
+epochs = np.arange(1, 51)
+
+# Training loss: exponential decay with noise (continues decreasing)
+train_loss = 2.5 * np.exp(-0.08 * epochs) + 0.15 + np.random.normal(0, 0.02, len(epochs))
+
+# Validation loss: decay then overfitting (U-shape after minimum)
+val_base = 2.5 * np.exp(-0.07 * epochs) + 0.35
+val_loss = val_base + np.random.normal(0, 0.025, len(epochs))
+# Add overfitting: loss increases after epoch 25
+val_loss[25:] = val_loss[25:] + np.linspace(0, 0.35, 25)
+
+# Find minimum validation loss epoch for annotation
+min_val_epoch = epochs[np.argmin(val_loss)]
+min_val_loss = np.min(val_loss)
+
+# Create DataFrame in long format for Altair
+df = pd.DataFrame(
+    {
+        "Epoch": np.tile(epochs, 2),
+        "Loss": np.concatenate([train_loss, val_loss]),
+        "Type": ["Training Loss"] * len(epochs) + ["Validation Loss"] * len(epochs),
+    }
+)
+
+# Point for minimum validation loss annotation
+min_point_df = pd.DataFrame({"Epoch": [min_val_epoch], "Loss": [min_val_loss], "Type": ["Optimal Stopping Point"]})
+
+# Base line chart
+lines = (
+    alt.Chart(df)
+    .mark_line(strokeWidth=3)
+    .encode(
+        x=alt.X("Epoch:Q", title="Epoch", axis=alt.Axis(labelFontSize=18, titleFontSize=22)),
+        y=alt.Y("Loss:Q", title="Cross-Entropy Loss", axis=alt.Axis(labelFontSize=18, titleFontSize=22)),
+        color=alt.Color(
+            "Type:N",
+            scale=alt.Scale(domain=["Training Loss", "Validation Loss"], range=["#306998", "#FFD43B"]),
+            legend=alt.Legend(title="Curve Type", labelFontSize=16, titleFontSize=18),
+        ),
+    )
+)
+
+# Add points on lines for visibility
+points = (
+    alt.Chart(df)
+    .mark_point(size=60, filled=True)
+    .encode(
+        x="Epoch:Q",
+        y="Loss:Q",
+        color=alt.Color(
+            "Type:N",
+            scale=alt.Scale(domain=["Training Loss", "Validation Loss"], range=["#306998", "#FFD43B"]),
+            legend=None,
+        ),
+    )
+)
+
+# Annotation for minimum validation loss
+min_marker = (
+    alt.Chart(min_point_df)
+    .mark_point(size=300, shape="diamond", filled=True, color="#E63946")
+    .encode(x="Epoch:Q", y="Loss:Q")
+)
+
+# Text annotation for optimal stopping point
+min_text = (
+    alt.Chart(min_point_df)
+    .mark_text(align="left", dx=12, dy=-10, fontSize=16, fontWeight="bold", color="#E63946")
+    .encode(x="Epoch:Q", y="Loss:Q", text=alt.value(f"Min Val Loss (Epoch {min_val_epoch})"))
+)
+
+# Combine all layers
+chart = (
+    (lines + points + min_marker + min_text)
+    .properties(
+        width=1600,
+        height=900,
+        title=alt.Title("line-loss-training · altair · pyplots.ai", fontSize=28, anchor="middle"),
+    )
+    .configure_axis(labelFontSize=18, titleFontSize=22, gridOpacity=0.3)
+    .configure_legend(labelFontSize=16, titleFontSize=18)
+    .configure_view(strokeWidth=0)
+)
+
+# Save as PNG (4800 x 2700 with scale_factor=3)
+chart.save("plot.png", scale_factor=3.0)
+
+# Save as HTML for interactivity
+chart.save("plot.html")
diff --git a/plots/line-loss-training/metadata/altair.yaml b/plots/line-loss-training/metadata/altair.yaml
new file mode 100644
index 0000000000..e5e04d1b7e
--- /dev/null
+++ b/plots/line-loss-training/metadata/altair.yaml
@@ -0,0 +1,28 @@
+library: altair
+specification_id: line-loss-training
+created: '2025-12-31T00:11:50Z'
+updated: '2025-12-31T00:16:02Z'
+generated_by: claude-opus-4-5-20251101
+workflow_run: 20608675031
+issue: 2860
+python_version: 3.13.11
+library_version: 6.0.0
+preview_url: https://storage.googleapis.com/pyplots-images/plots/line-loss-training/altair/plot.png
+preview_thumb: https://storage.googleapis.com/pyplots-images/plots/line-loss-training/altair/plot_thumb.png
+preview_html: https://storage.googleapis.com/pyplots-images/plots/line-loss-training/altair/plot.html
+quality_score: 91
+review:
+  strengths:
+  - Excellent implementation of overfitting visualization with validation loss diverging
+    from training loss after epoch 25
+  - Clean use of Altair's layered composition with separate chart objects for lines,
+    points, and annotations
+  - Red diamond marker with annotation clearly indicates the optimal stopping point
+    at minimum validation loss
+  - Colorblind-safe color scheme with strong contrast between blue training and yellow
+    validation curves
+  - Proper title format and axis labeling including loss function type as specified
+  weaknesses:
+  - Legend title "Curve Type" is generic; could be more descriptive
+  - Points on every epoch (50 points per curve) add visual noise; could use opacity
+    or fewer markers