Skip to content

Fix low max tokens value for Anthropic when it supports thinking #564

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
haiphucnguyen merged 1 commit into from
Jul 27, 2026
Merged

Fix low max tokens value for Anthropic when it supports thinking #564

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions gradle/libs.versions.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,9 +1,9 @@
[versions]
graalvm-plugin = "0.11.5"
jline = "3.30.6"
langchain4j = "1.17.1"
langchain4j-extra = "1.17.1-beta27"
langchain4j-jvector = "1.17.0-beta27"
langchain4j = "1.18.0"
langchain4j-extra = "1.18.0-beta28"
langchain4j-jvector = "1.18.0-beta28"
commonmark = "0.29.0"
kotlinxSerializationJson = "1.11.0"
kotlinxCoroutines = "1.11.0"
Expand Down
6 changes: 5 additions & 1 deletion shared/src/main/kotlin/io/askimo/core/providers/anthropic/AnthropicModelFactory.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -183,18 +183,22 @@ class AnthropicModelFactory : ChatModelFactory<AnthropicSettings> {
.cacheTools(true)
.logger(log)
.logRequests(log.isDebugEnabled)
.logResponses(log.isTraceEnabled)
.logResponses(log.isDebugEnabled)
.listeners(listOf(TelemetryChatModelListener(telemetry, ANTHROPIC.name.lowercase())))
.apply {
if (supportsThinking) {
if (reasoningLevel.isEnabled) {
thinkingType("adaptive")
sendThinking(true)
returnThinking(true)
.maxTokens(32_000)
} else {
// OFF — explicitly disable extended thinking
thinkingType("disabled")
maxTokens(8192)
}
} else {
maxTokens(8192)
}
}
.build()
Expand Down
7 changes: 2 additions & 5 deletions shared/src/main/kotlin/io/askimo/core/providers/gemini/GeminiModelFactory.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,10 +91,8 @@ class GeminiModelFactory : ChatModelFactory<GeminiSettings> {
) {
ModelCapabilitiesCache.setToolSupport(GEMINI, settings.defaultModel, true)
val modelName = settings.defaultModel
val capturedModel = streamingModel
val capturedMode = executionMode
CoroutineScope(Dispatchers.IO + SupervisorJob()).launch {
val supportsTools = probeToolSupport(modelName, capturedModel, capturedMode)
val supportsTools = probeToolSupport(modelName, streamingModel, executionMode)
ModelCapabilitiesCache.setToolSupport(GEMINI, modelName, supportsTools)
}
}
Expand All @@ -106,9 +104,8 @@ class GeminiModelFactory : ChatModelFactory<GeminiSettings> {
if (!ModelCapabilitiesCache.hasTestedImageSupport(GEMINI, settings.defaultModel)) {
ModelCapabilitiesCache.setImageSupport(GEMINI, settings.defaultModel, false)
val modelName = settings.defaultModel
val capturedModel = streamingModel
CoroutineScope(Dispatchers.IO + SupervisorJob()).launch {
val supportsImage = probeImageCapability(GEMINI, modelName, capturedModel)
val supportsImage = probeImageCapability(GEMINI, modelName, streamingModel)
ModelCapabilitiesCache.setImageSupport(GEMINI, modelName, supportsImage)
}
}
Expand Down