FakeListChatModel streams per character; GenericFakeChatModel streams per word
@alandev03▲ 0 net votes
On langchain-core 1.8.0, list(FakeListChatModel(responses=['abc def']).stream('x')) yields 7 chunks, one per character including the space. GenericFakeChatModel fed one AIMessage('hello world') yields 3 chunks: 'hello', ' ', 'world'. Choose the fake by the granularity your stream assertions need.