Surface ModelBilling.tokenPrices on public SDK types

dotnet/src/Types.cs

@@ -3308,6 +3308,70 @@ public sealed class ModelBilling
     /// </summary>
     [JsonPropertyName("multiplier")]
     public double? Multiplier { get; set; }
+
+    /// <summary>
+    /// Token-level pricing information for this model.
+    /// </summary>
+    [JsonPropertyName("tokenPrices")]
+    public ModelBillingTokenPrices? TokenPrices { get; set; }
+}
+
+/// <summary>
+/// Token-level pricing information for a model
+/// </summary>
+public sealed class ModelBillingTokenPrices
+{
+    /// <summary>AI Credits cost per billing batch of input tokens.</summary>
+    [JsonPropertyName("inputPrice")]
+    public double? InputPrice { get; set; }
+
+    /// <summary>AI Credits cost per billing batch of output tokens.</summary>
+    [JsonPropertyName("outputPrice")]
+    public double? OutputPrice { get; set; }
+
+    /// <summary>AI Credits cost per billing batch of cached tokens.</summary>
+    [JsonPropertyName("cachePrice")]
+    public double? CachePrice { get; set; }
+
+    /// <summary>Number of tokens per standard billing batch.</summary>
+    [JsonPropertyName("batchSize")]
+    public int? BatchSize { get; set; }
+
+    /// <summary>
+    /// Prompt token budget (max_prompt_tokens) for the default tier. The total
+    /// context window is this value plus the model's max_output_tokens.
+    /// </summary>
+    [JsonPropertyName("contextMax")]
+    public int? ContextMax { get; set; }
+
+    /// <summary>Long context tier pricing (available for models with extended context windows).</summary>
+    [JsonPropertyName("longContext")]
+    public ModelBillingTokenPricesLongContext? LongContext { get; set; }
+}
+
+/// <summary>
+/// Long context tier pricing (available for models with extended context windows)
+/// </summary>
+public sealed class ModelBillingTokenPricesLongContext
+{
+    /// <summary>AI Credits cost per billing batch of input tokens.</summary>
+    [JsonPropertyName("inputPrice")]
+    public double? InputPrice { get; set; }
+
+    /// <summary>AI Credits cost per billing batch of output tokens.</summary>
+    [JsonPropertyName("outputPrice")]
+    public double? OutputPrice { get; set; }
+
+    /// <summary>AI Credits cost per billing batch of cached tokens.</summary>
+    [JsonPropertyName("cachePrice")]
+    public double? CachePrice { get; set; }
+
+    /// <summary>
+    /// Prompt token budget (max_prompt_tokens) for the long context tier. The total
+    /// context window is this value plus the model's max_output_tokens.
+    /// </summary>
+    [JsonPropertyName("contextMax")]
+    public int? ContextMax { get; set; }
 }
 
 /// <summary>
@@ -3509,6 +3573,8 @@ public sealed class SystemMessageTransformRpcResponse
 [JsonSerializable(typeof(McpServerConfig))]
 [JsonSerializable(typeof(MessageOptions))]
 [JsonSerializable(typeof(ModelBilling))]
+[JsonSerializable(typeof(ModelBillingTokenPrices))]
+[JsonSerializable(typeof(ModelBillingTokenPricesLongContext))]
 [JsonSerializable(typeof(ModelCapabilities))]
 [JsonSerializable(typeof(ModelCapabilitiesOverride))]
 [JsonSerializable(typeof(ModelInfo))]

dotnet/test/Unit/SerializationTests.cs

@@ -50,6 +50,56 @@ public void ProviderConfig_CanSerializeHeaders_WithSdkOptions()
         Assert.Equal(4096, deserialized.MaxOutputTokens);
     }
 
+    [Fact]
+    public void ModelBilling_CanSerializeTokenPrices_WithSdkOptions()
+    {
+        var options = GetSerializerOptions();
+        var original = new ModelBilling
+        {
+            Multiplier = 1.5,
+            TokenPrices = new ModelBillingTokenPrices
+            {
+                InputPrice = 2.0,
+                OutputPrice = 8.0,
+                CachePrice = 0.5,
+                BatchSize = 1_000_000,
+                ContextMax = 128_000,
+                LongContext = new ModelBillingTokenPricesLongContext
+                {
+                    InputPrice = 4.0,
+                    OutputPrice = 16.0,
+                    CachePrice = 1.0,
+                    ContextMax = 1_000_000
+                }
+            }
+        };
+
+        var json = JsonSerializer.Serialize(original, options);
+        using var document = JsonDocument.Parse(json);
+        var root = document.RootElement;
+        Assert.Equal(1.5, root.GetProperty("multiplier").GetDouble());
+        var tokenPrices = root.GetProperty("tokenPrices");
+        Assert.Equal(2.0, tokenPrices.GetProperty("inputPrice").GetDouble());
+        Assert.Equal(8.0, tokenPrices.GetProperty("outputPrice").GetDouble());
+        Assert.Equal(0.5, tokenPrices.GetProperty("cachePrice").GetDouble());
+        Assert.Equal(1_000_000, tokenPrices.GetProperty("batchSize").GetInt32());
+        Assert.Equal(128_000, tokenPrices.GetProperty("contextMax").GetInt32());
+        var longContext = tokenPrices.GetProperty("longContext");
+        Assert.Equal(4.0, longContext.GetProperty("inputPrice").GetDouble());
+        Assert.Equal(1_000_000, longContext.GetProperty("contextMax").GetInt32());
+
+        var deserialized = JsonSerializer.Deserialize<ModelBilling>(json, options);
+        Assert.NotNull(deserialized);
+        Assert.Equal(1.5, deserialized.Multiplier);
+        Assert.NotNull(deserialized.TokenPrices);
+        Assert.Equal(2.0, deserialized.TokenPrices.InputPrice);
+        Assert.Equal(1_000_000, deserialized.TokenPrices.BatchSize);
+        Assert.Equal(128_000, deserialized.TokenPrices.ContextMax);
+        Assert.NotNull(deserialized.TokenPrices.LongContext);
+        Assert.Equal(16.0, deserialized.TokenPrices.LongContext.OutputPrice);
+        Assert.Equal(1_000_000, deserialized.TokenPrices.LongContext.ContextMax);
+    }
+
     [Fact]
     public void MessageOptions_CanSerializeRequestHeaders_WithSdkOptions()
     {

go/client_test.go

@@ -892,6 +892,75 @@ func TestListModelsWithCustomHandler(t *testing.T) {
 	}
 }
 
+func TestModelBillingTokenPricesJSON(t *testing.T) {
+	wire := `{
+		"multiplier": 1.5,
+		"tokenPrices": {
+			"inputPrice": 2.0,
+			"outputPrice": 8.0,
+			"cachePrice": 0.5,
+			"batchSize": 1000000,
+			"contextMax": 128000,
+			"longContext": {
+				"inputPrice": 4.0,
+				"outputPrice": 16.0,
+				"cachePrice": 1.0,
+				"contextMax": 1000000
+			}
+		}
+	}`
+
+	var billing ModelBilling
+	if err := json.Unmarshal([]byte(wire), &billing); err != nil {
+		t.Fatalf("unmarshal failed: %v", err)
+	}
+
+	if billing.TokenPrices == nil {
+		t.Fatal("expected TokenPrices to be set")
+	}
+	tp := billing.TokenPrices
+	if tp.InputPrice == nil || *tp.InputPrice != 2.0 {
+		t.Errorf("unexpected InputPrice: %v", tp.InputPrice)
+	}
+	if tp.OutputPrice == nil || *tp.OutputPrice != 8.0 {
+		t.Errorf("unexpected OutputPrice: %v", tp.OutputPrice)
+	}
+	if tp.CachePrice == nil || *tp.CachePrice != 0.5 {
+		t.Errorf("unexpected CachePrice: %v", tp.CachePrice)
+	}
+	if tp.BatchSize == nil || *tp.BatchSize != 1000000 {
+		t.Errorf("unexpected BatchSize: %v", tp.BatchSize)
+	}
+	if tp.ContextMax == nil || *tp.ContextMax != 128000 {
+		t.Errorf("unexpected ContextMax: %v", tp.ContextMax)
+	}
+	if tp.LongContext == nil {
+		t.Fatal("expected LongContext to be set")
+	}
+	lc := tp.LongContext
+	if lc.InputPrice == nil || *lc.InputPrice != 4.0 {
+		t.Errorf("unexpected LongContext.InputPrice: %v", lc.InputPrice)
+	}
+	if lc.ContextMax == nil || *lc.ContextMax != 1000000 {
+		t.Errorf("unexpected LongContext.ContextMax: %v", lc.ContextMax)
+	}
+
+	// Round-trip back to JSON and ensure the nested structure survives.
+	out, err := json.Marshal(billing)
+	if err != nil {
+		t.Fatalf("marshal failed: %v", err)
+	}
+	var reparsed ModelBilling
+	if err := json.Unmarshal(out, &reparsed); err != nil {
+		t.Fatalf("re-unmarshal failed: %v", err)
+	}
+	if reparsed.TokenPrices == nil || reparsed.TokenPrices.LongContext == nil ||
+		reparsed.TokenPrices.LongContext.ContextMax == nil ||
+		*reparsed.TokenPrices.LongContext.ContextMax != 1000000 {
+		t.Errorf("round-trip lost token price data: %s", out)
+	}
+}
+
 func TestListModelsHandlerCachesResults(t *testing.T) {
 	customModels := []ModelInfo{
 		{

go/types.go

@@ -1599,7 +1599,40 @@ type ModelPolicy struct {
 
 // ModelBilling contains model billing information
 type ModelBilling struct {
-	Multiplier *float64 `json:"multiplier,omitempty"`
+	Multiplier  *float64                 `json:"multiplier,omitempty"`
+	TokenPrices *ModelBillingTokenPrices `json:"tokenPrices,omitempty"`
+}
+
+// ModelBillingTokenPrices contains token-level pricing information for a model
+type ModelBillingTokenPrices struct {
+	// InputPrice is the AI Credits cost per billing batch of input tokens
+	InputPrice *float64 `json:"inputPrice,omitempty"`
+	// OutputPrice is the AI Credits cost per billing batch of output tokens
+	OutputPrice *float64 `json:"outputPrice,omitempty"`
+	// CachePrice is the AI Credits cost per billing batch of cached tokens
+	CachePrice *float64 `json:"cachePrice,omitempty"`
+	// BatchSize is the number of tokens per standard billing batch
+	BatchSize *int `json:"batchSize,omitempty"`
+	// ContextMax is the prompt token budget (max_prompt_tokens) for the default
+	// tier. The total context window is this value plus the model's max_output_tokens.
+	ContextMax *int `json:"contextMax,omitempty"`
+	// LongContext is the long context tier pricing (available for models with
+	// extended context windows)
+	LongContext *ModelBillingTokenPricesLongContext `json:"longContext,omitempty"`
+}
+
+// ModelBillingTokenPricesLongContext contains long context tier pricing
+// (available for models with extended context windows)
+type ModelBillingTokenPricesLongContext struct {
+	// InputPrice is the AI Credits cost per billing batch of input tokens
+	InputPrice *float64 `json:"inputPrice,omitempty"`
+	// OutputPrice is the AI Credits cost per billing batch of output tokens
+	OutputPrice *float64 `json:"outputPrice,omitempty"`
+	// CachePrice is the AI Credits cost per billing batch of cached tokens
+	CachePrice *float64 `json:"cachePrice,omitempty"`
+	// ContextMax is the prompt token budget (max_prompt_tokens) for the long
+	// context tier. The total context window is this value plus the model's max_output_tokens.
+	ContextMax *int `json:"contextMax,omitempty"`
 }
 
 // ModelInfo contains information about an available model

java/src/main/java/com/github/copilot/rpc/ModelBilling.java

@@ -16,14 +16,26 @@
 public class ModelBilling {
 
     @JsonProperty("multiplier")
-    private double multiplier;
+    private Double multiplier;
 
-    public double getMultiplier() {
+    @JsonProperty("tokenPrices")
+    private ModelBillingTokenPrices tokenPrices;
+
+    public Double getMultiplier() {
         return multiplier;
     }
 
-    public ModelBilling setMultiplier(double multiplier) {
+    public ModelBilling setMultiplier(Double multiplier) {
         this.multiplier = multiplier;
         return this;
     }
+
+    public ModelBillingTokenPrices getTokenPrices() {
+        return tokenPrices;
+    }
+
+    public ModelBilling setTokenPrices(ModelBillingTokenPrices tokenPrices) {
+        this.tokenPrices = tokenPrices;
+        return this;
+    }
 }

java/src/main/java/com/github/copilot/rpc/ModelBillingTokenPrices.java

@@ -0,0 +1,109 @@
+/*---------------------------------------------------------------------------------------------
+ *  Copyright (c) Microsoft Corporation. All rights reserved.
+ *--------------------------------------------------------------------------------------------*/
+
+package com.github.copilot.rpc;
+
+import com.fasterxml.jackson.annotation.JsonIgnoreProperties;
+import com.fasterxml.jackson.annotation.JsonProperty;
+
+/**
+ * Token-level pricing information for a model.
+ *
+ * @since 1.0.2
+ */
+@JsonIgnoreProperties(ignoreUnknown = true)
+public class ModelBillingTokenPrices {
+
+    /**
+     * AI Credits cost per billing batch of input tokens.
+     */
+    @JsonProperty("inputPrice")
+    private Double inputPrice;
+
+    /**
+     * AI Credits cost per billing batch of output tokens.
+     */
+    @JsonProperty("outputPrice")
+    private Double outputPrice;
+
+    /**
+     * AI Credits cost per billing batch of cached tokens.
+     */
+    @JsonProperty("cachePrice")
+    private Double cachePrice;
+
+    /**
+     * Number of tokens per standard billing batch.
+     */
+    @JsonProperty("batchSize")
+    private Integer batchSize;
+
+    /**
+     * Prompt token budget (max_prompt_tokens) for the default tier. The total
+     * context window is this value plus the model's max_output_tokens.
+     */
+    @JsonProperty("contextMax")
+    private Integer contextMax;
+
+    /**
+     * Long context tier pricing (available for models with extended context
+     * windows).
+     */
+    @JsonProperty("longContext")
+    private ModelBillingTokenPricesLongContext longContext;
+
+    public Double getInputPrice() {
+        return inputPrice;
+    }
+
+    public ModelBillingTokenPrices setInputPrice(Double inputPrice) {
+        this.inputPrice = inputPrice;
+        return this;
+    }
+
+    public Double getOutputPrice() {
+        return outputPrice;
+    }
+
+    public ModelBillingTokenPrices setOutputPrice(Double outputPrice) {
+        this.outputPrice = outputPrice;
+        return this;
+    }
+
+    public Double getCachePrice() {
+        return cachePrice;
+    }
+
+    public ModelBillingTokenPrices setCachePrice(Double cachePrice) {
+        this.cachePrice = cachePrice;
+        return this;
+    }
+
+    public Integer getBatchSize() {
+        return batchSize;
+    }
+
+    public ModelBillingTokenPrices setBatchSize(Integer batchSize) {
+        this.batchSize = batchSize;
+        return this;
+    }
+
+    public Integer getContextMax() {
+        return contextMax;
+    }
+
+    public ModelBillingTokenPrices setContextMax(Integer contextMax) {
+        this.contextMax = contextMax;
+        return this;
+    }
+
+    public ModelBillingTokenPricesLongContext getLongContext() {
+        return longContext;
+    }
+
+    public ModelBillingTokenPrices setLongContext(ModelBillingTokenPricesLongContext longContext) {
+        this.longContext = longContext;
+        return this;
+    }
+}