OLMo 3 32B Think is Allen Institute for AI's fully open 32B reasoning model, released with open weights, training data, and code so its chain-of-thought can be inspected and traced back to training decisions. It is tuned for math and coding reasoning via long chain-of-thought, though it lacks tool use or vision support, making it best suited for transparent, auditable reasoning tasks rather than agentic workflows.
Mức chi phí
0.1x
Ngữ cảnh
66K
Phát hành
21 thg 11, 2025
Đầu vào
Text
Đầu ra
Text
Hỗ trợ
Suy luậnGọi công cụĐầu ra có cấu trúc