Moonshot Releases Kimi K2
teortaxesTex · x · 2026-07-12
Moonshot has released Kimi K2, its first new frontier model in a year. The original post highlights it as an open-source agentic model with 1T total / 32B active MoE parameters. It shows strong performance on open model benchmarks like SWE Bench Verified, Tau2, and AceBench, making it particularly suited for coding and agentic tasks.
API and open-source resources provided:
- API is now open
- Pricing: Input $0.15 / million tokens (cache hit), $0.60 / million tokens (cache miss); Output $2.50 / million tokens
- Includes a tech blog, weights & code, GitHub, and a trial link
The post also notes that multimodal and thought mode are currently unsupported.
More from Models
- Nanbeige4.3-33B benchmarked: 59 tok/s single stream, ~1,060 tok/s across 64 concurrent runs — MaziyarPanahi · 2026-10-26
- Only 2 of 25 full ChatGPT rewrites flagged 100% AI by detectors — paulnovosad · 2026-10-26
- Qwen 4 Reportedly Spotted in the Arena, Hints at Alibaba's Next-Gen Model Testing — AIFlow_ML · 2026-10-26
- Nathan Lambert's first open models lecture: distillation, risks and the ecosystem explained — natolambert · 2026-10-26
- Nanbeige 4.3-33B flops real-world test: zero compilable output vs Qwen — julianharris · 2026-10-26
- Claude's safety classifier blocks SS7 defense research, pushing a security dev to an abliterated GLM — l4rz · 2026-10-26