证据：GPT-5.6 Luna model page (checked 2026-10-23): model id gpt-5.6-luna; inputs text and image, output text; 1,050,000 context window; maximum input 922,000 tokens; 128,000 max output tokens; February 16, 2026 knowledge cutoff; reasoning token support. Text pricing: input $0.2, cached input $0.02, output $1.2 per 1M tokens.
cached_input_price_usd_per_mtok
公开事实
0.02 USD/MTok
高置信度，已绑定来源与证据。
条件：Cache writes are billed at 1.25x the uncached input token rate.
证据：GPT-5.6 Luna model page (checked 2026-10-23): model id gpt-5.6-luna; inputs text and image, output text; 1,050,000 context window; maximum input 922,000 tokens; 128,000 max output tokens; February 16, 2026 knowledge cutoff; reasoning token support. Text pricing: input $0.2, cached input $0.02, output $1.2 per 1M tokens.
capability_tags
公开事实
agent、code、image_understanding、reasoning
高置信度，已绑定来源与证据。
条件：Capabilities inferred from the model page statements: cost-sensitive high-volume workloads, reasoning token support, image input and tool calling.
证据：GPT-5.6 Luna model page (checked 2026-10-23): model id gpt-5.6-luna; inputs text and image, output text; 1,050,000 context window; maximum input 922,000 tokens; 128,000 max output tokens; February 16, 2026 knowledge cutoff; reasoning token support. Text pricing: input $0.2, cached input $0.02, output $1.2 per 1M tokens.
证据：GPT-5.6 Luna model page (checked 2026-10-23): model id gpt-5.6-luna; inputs text and image, output text; 1,050,000 context window; maximum input 922,000 tokens; 128,000 max output tokens; February 16, 2026 knowledge cutoff; reasoning token support. Text pricing: input $0.2, cached input $0.02, output $1.2 per 1M tokens.
input_price_usd_per_mtok
公开事实
0.2 USD/MTok
高置信度，已绑定来源与证据。
条件：Promotional pricing in effect after the July 30, 2026 reduction; prompts above 272K input tokens are priced at 2x input and 1.5x output.
证据：GPT-5.6 Luna model page (checked 2026-10-23): model id gpt-5.6-luna; inputs text and image, output text; 1,050,000 context window; maximum input 922,000 tokens; 128,000 max output tokens; February 16, 2026 knowledge cutoff; reasoning token support. Text pricing: input $0.2, cached input $0.02, output $1.2 per 1M tokens.
证据：OpenAI changelog (checked 2026-10-23): on July 9, 2026 OpenAI released the GPT-5.6 model family including GPT-5.6 Luna for efficient high-volume workloads; a July 30, 2026 entry states GPT-5.6 Luna costs 80% less; later entries list gpt-5.6-luna as a current model.
证据：GPT-5.6 Luna model page (checked 2026-10-23): model id gpt-5.6-luna; inputs text and image, output text; 1,050,000 context window; maximum input 922,000 tokens; 128,000 max output tokens; February 16, 2026 knowledge cutoff; reasoning token support. Text pricing: input $0.2, cached input $0.02, output $1.2 per 1M tokens.
证据：OpenAI changelog (checked 2026-10-23): on July 9, 2026 OpenAI released the GPT-5.6 model family including GPT-5.6 Luna for efficient high-volume workloads; a July 30, 2026 entry states GPT-5.6 Luna costs 80% less; later entries list gpt-5.6-luna as a current model.