Unknown Error code 1032 from Apple Intelligence
20:19 10 May 2026

I'm using Apple intelligence to power suggestions to user of my app. It's a great feature but I'm baffled about this error.

Domain=FoundationModels.LanguageModelSession.GenerationError Code=-1 "(null)" UserInfo={NSMultipleUnderlyingErrorsKey=(
    "Error Domain=ModelManagerServices.ModelManagerError Code=1032 \"(null)\" UserInfo={NSMultipleUnderlyingErrorsKey=(\n)}"
)} 

My code uses the "stream" feature and @Generable structures to give the user a great experience.

    func generateIdeas(name: String, hero: String, creativity: Creativity, numberOfIdeas: Int) throws -> LanguageModelSession.ResponseStream<[ColorChip]> {

        try self.checkAvailability()
        
        let prompt = """
        Please generate at least \(numberOfIdeas) different and unique primary colors for the users idea which they are calling -> \'\(name)\' and pitch -> \'\(hero)\'
        """

        let assistant = colorIdeasModelSession.streamResponse(to: prompt,
                                                            generating: [ColorChip].self,
                                                            includeSchemaInPrompt: true,
                                                            options: creativity.generationOptions)
        
        return assistant
    }

You can see I check the availability and so there is no doubt here is that code:

    private func checkAvailability(forModel: SystemLanguageModel = .default) throws {
        switch forModel.availability {
        case .available:
            return
        case .unavailable(let reason):
            throw AvailabilityError.unavailable(reason)
        }
    }

I also have Apple intelligence turned on and my system language settings matching my Siri language.

enter image description here

Now, when I google this error Google has its AI surface this article from someone with a similar issue. I made sure my languages are correct.

To make the issue more infuriating the error does not actually get thrown until I start listening for the stream in my calling code.

    @MainActor
    func generate() async {
        
        self.generatingIdeas = []
        focusField = nil
        
        do {
            
            withAnimation(.snappy) {
                generationStep = .streaming
            }
            
            let ideaStream = try agent.generateIdeas(name: self.projectName, hero: self.projectPitch, creativity: Creativity.allCases.randomElement() ?? .medium, numberOfIdeas: 5)
            
            //Error gets throws from here.
            for try await partialResponse in ideaStream {
                self.generatingIdeas = partialResponse.content
            }
            
            withAnimation(.snappy) {
                generationStep = .finished
            }
            
            generationText = "Generate again"
        }catch {
            generationStep = .error(error)
        }
    }

What gives? This happens on 2 of my three devices yet I can't find anything on the internet about it? I have filed it in their Feedback Assistant system.

swift xcode apple-intelligence foundation-models