Meituan launches LongCat 2.0: 1.6T parameter MoE model with 1M+ context window at $0.30 per 1M input tokens
Meituan has released LongCat 2.0, a sparse mixture-of-experts language model with 48 billion active parameters out of 1.6 trillion total. The model features a 1,049,000 token context window and costs $0.30 per 1M input tokens and $1.20 per 1M output tokens.
LongCat 2.0 — Quick Specs
Meituan launches LongCat 2.0: 1.6T parameter MoE model with 1M+ context window
Meituan has released LongCat 2.0, a sparse mixture-of-experts (MoE) language model with 48 billion active parameters out of 1.6 trillion total parameters. The model is now available through OpenRouter's API.
Model specifications
LongCat 2.0 features a 1,049,000 token context window, placing it among the largest context windows available in production models. The sparse MoE architecture activates only 48 billion parameters per inference call while maintaining access to the full 1.6 trillion parameter base.
Pricing is set at $0.30 per 1 million input tokens and $1.20 per 1 million output tokens, making it competitively priced for long-context applications.
Target use cases
According to Meituan, LongCat 2.0 is designed for:
- Coding tasks and repository-level code changes
- Long-horizon problem solving
- Agentic workflows requiring extended context
The model operates as a text-to-text model and is accessed via the model ID meituan/longcat-2.0 on OpenRouter.
Technical architecture
The sparse MoE design allows LongCat 2.0 to maintain efficiency despite its massive total parameter count. By activating only 3% of its parameters (48B out of 1.6T) during inference, the model aims to balance computational cost with capability.
No benchmark scores have been publicly disclosed at this time. Training data cutoff date and additional technical details remain unannounced.
What this means
LongCat 2.0's million-token context window positions it for applications requiring extensive code repository analysis or long-document processing. The sparse MoE architecture suggests Meituan is following the industry trend of scaling total parameters while keeping active parameters manageable. At $0.30 per 1M input tokens, it's priced below several competitors offering similar context lengths, though real-world performance data will determine whether the pricing advantage translates to practical value. The model's availability through OpenRouter provides immediate API access without requiring direct integration with Meituan's infrastructure.
Related Articles
Moonshot AI Releases Kimi K3: 2.8T Parameter Open Model at $3/$15 Per Million Tokens
Moonshot AI has released Kimi K3, a 2.8 trillion parameter model with 1 million token context window and native multimodal input. The model ranks #1 in Frontend Code Arena and #9 in Text Arena, with pricing at $3 per million input tokens and $15 per million output tokens—comparable to Claude Sonnet 5 pricing while delivering performance the company claims is near Claude Opus 4.8 and GPT-5.5.
Thinking Machines releases Inkling: 975B-parameter MoE model with Apache 2.0 license, first major US open-weight multimo
Thinking Machines Lab released Inkling, a mixture-of-experts model with 975B total parameters and 41B active parameters, trained on 45 trillion tokens across text, images, audio, and video. The Apache 2.0-licensed model supports up to 1M context and debuts alongside Inkling-Small (276B-A12B), marking what observers call the strongest US-based open-weight release to date.
Moonshot AI's Kimi K3 ranks #2 globally, will release 2.8T parameter weights July 27
Moonshot AI released Kimi K3 on July 16, 2026, a 2.8 trillion parameter mixture-of-experts model that ranks #2 on the Vals AI index and #3 on Artificial Analysis's Intelligence Index. The company will release the model's weights on July 27, making it the strongest open-weight model to date, surpassing all previous open releases including DeepSeek R1.
Moonshot AI's Kimi k3 claims top performance among Chinese models with 1M token context
Moonshot AI has released Kimi k3, positioning it as China's leading AI model. The company claims the model features a 1 million token context window and improved reasoning capabilities, though independent benchmarks are not yet available.
Comments
Loading...