Thinking Machines Lab, an AI startup founded by former OpenAI CTO Mira Murati, has released Inkling 2 Small. Third-party tester Artificial Analysis says the model could give Chinese open-weight models a run for their lead. The company said it will release the model's weights next week.
Inkling 2 Small has 594 billion parameters. It accepts image inputs but outputs only text. Thinking Machines said the model is stronger at coding and easier to use inside real agent loops.
Artificial Analysis scored the model 31 on its Intelligence Index, which combines benchmarks covering coding, science and general categories. Artificial Analysis said that score is above average for its class and ties with Gemini 3.8 Flash. The index's top scorers are Gemini 4 Carbon at 60 and Claude Opus 5.5 at 59.
Artificial Analysis said that if the weights are released as planned, Inkling 2 Small would be the highest-scoring open-weight model developed outside China based on current rankings.
The firm also reported drawbacks. It called the model "notably slow per task and very verbose." It described the price of $0.30 per million input tokens as "somewhat expensive," against a median of $0.24 for its class.
Artificial Analysis said the model improved on its predecessor, with a 41% gain in automation capabilities and a 20% gain in coding abilities. On the AA-Omniscience v1.1 test, the hallucination rate fell from 61% to 27%, accuracy rose from 32% to 34%, and the attempt rate fell from 77% to 53%.