Model Detail
Kimi-K2-Instruct
—Kimi-K2-Instruct is a code generation model with 513.2B parameters released by moonshotai. The model is registered under the text-generation pipeline tag on Hugging Face, and supports text->text inputs, distributed under a other license.
Kimi-K2-Instruct has been evaluated across multiple task suites. On task-specific evaluations the model scores 65.4% resolved on SWE-Bench.
Kimi-K2-Instruct is priced at $0.5/M input tokens and $2/M output tokens. Operationally the model offers a 131K-token context window, which matters when sizing it for prompt-heavy or latency-sensitive workloads. At this input rate the model sits in the commodity tier and is suitable for high-volume workloads where per-call cost dominates the decision.
Kimi-K2-Instruct ships with 513.2B parameters. The published knowledge cutoff is 2024-12-31, so newer events will not be reflected in zero-shot answers without retrieval. Total weight footprint is approximately 1026.4 GB, which is the relevant figure when planning local-inference VRAM. Distribution is governed by the other license — review the exact terms before commercial deployment.
Kimi-K2-Instruct is best fit for code completion, repository-scale Q&A, and pair-programming integrations, and high-volume batch jobs where per-call cost dominates the budget. It is a less obvious choice for one-shot generation of security-critical code without review. Treat this as a starting matrix rather than a benchmark verdict — the right deployment usually depends on the specific evaluation suite that mirrors your workload.
Complex KDA: Understanding and Enhancing the Expressivity of Kimi Delta Attention
arXiv:2609.24797v1 Announce Type: new Abstract: Linear RNNs based on the delta-rule enable efficient sequence modeling, but their linear updates with a low-rank correction constrain their expressivity. Prior work has shown that composing two delta-rule transitions in a single recurrent update can mo
Kimi-maker Moonshot AI targets $2B in annual revenue
While K3's usage figures have declined slightly in recent months, OpenRouter data currently shows as many as 300 billion tokens being generated each day by K3 models on the system.
Kimi K3: Open Frontier Intelligence
arXiv:2607.24653v1 Announce Type: cross Abstract: We introduce Kimi K3, a 2.8T parameter Mixture-of-Experts model with 104 billion activated parameters, native vision capabilities, and a 1-million-token context window. Kimi K3 is built on Kimi Delta Attention and Attention Residuals, which improve i
‘AI communism’, rogue models, and the why Kimi K3 spooked Wall Street
Chinese AI lab Moonshot’s open model Kimi went viral this week for reasons that had less to do with the model itself and more to do with how the U.S. AI industry reacted to it. Meanwhile, an unreleased OpenAI model wandered outside its test environment and ended up connected to a real security breac
Experts say exploiting Anthropic’s Fable isn’t how Kimi K3 got so good
"I don't think you get a model this strong and this quickly on the heels of Fable doing strictly distillation," one expert told TechCrunch.
Kimi: Threat or menace?
Chinese company Moonshot AI released a new version of its Kimi model this week, prompting concern about "full AI communism."
