Thinking Machines Lab has launched Inkling 2 Small, a 594 billion parameter Mixture-of-Experts model with 17 billion active parameters. This multimodal model accepts text and image inputs and supports a 1 million token context window. It demonstrates strong performance on various benchmarks, including SWE-Bench Pro V2 and PerceptionBench, and is available via the Tinker API. AI
IMPACT Sets a new benchmark for multimodal MoE models, potentially influencing agentic coding and vision tasks.
RANK_REASON Frontier-lab model release with system card and performance metrics. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Artificial Analysis Intelligence Index
- DeepSeek-V4.1-Flash-1017
- DesignArena Website
- EnterpriseBench-CoreCraft
- Inkling
- Inkling 2 Small
- JobBench
- Kimi k3
- NVIDIA GB300 NVL72
- PerceptionBench
- Perception Encoder
- Qwen3.5-4B
- SWE-bench Pro V2
- Thinking Machines Lab
- Tinker
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →