new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

moonshotai/options

3 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Moonshot AI Kimi K3 reasoning configuration

Kimi K3 always reasons and supports only the 'max' reasoning effort. Configure through provider options: providerOptions: { moonshotai: { reasoningEffort: 'max' } satisfies MoonshotAILanguageModelOptions }. The reasoning output is accessed via reasoningText from generateText response.

Moonshot AI thinking models configuration

Kimi thinking models like kimi-k2-thinking generate intermediate reasoning tokens. Configure with providerOptions: { moonshotai: { thinking: { type: 'enabled', budgetTokens: 2048 }, reasoningHistory: 'interleaved' } satisfies MoonshotAILanguageModelOptions }. The reasoning output streams through standard AI SDK reasoning parts and is accessed via reasoningText.

Moonshot AI provider options reference

Provider options for Moonshot AI language models: reasoningEffort (string, value 'max' for Kimi K3 reasoning); thinking (object with type: 'enabled' | 'disabled' and budgetTokens: number with minimum 1024 for thinking/reasoning models); reasoningHistory (string: 'disabled' removes reasoning from history, 'interleaved' includes reasoning between tool calls within a single turn, 'preserved' keeps all reasoning mapped to Moonshot's thinking.keep: 'all').

Give your agent this brain