Anthropic

由前 OpenAI 核心团队创立的 AI 安全公司，总部旧金山。Claude 系列以卓越的指令遵循、超长文本处理和代码生成能力著称，是开发者最信赖的编程助手之一。

Models

该厂商性价比之选

Claude Opus 5.5 (max)in $4 / out $20 · /1M tokens

所有模型按性价比排序

模型名称	上下文	智能指数
Claude Opus 5.5 (max)	1M	57.6
Claude Opus 5 (max)	1M	50.8
Claude Opus 5 (high)	1M	48.1
Claude Sonnet 5 (max)	1M	38.2
Claude Fable 5 (max)	1M	49.6
Claude Opus 4.7 (max)	1M	40.7
Claude Opus 4.8 (max)	1M	41.8
Claude Fable 5.1 (max)	1M	53.4
Claude Opus 5 (xhigh)	1M	49.7
Claude Fable 5.1 (xhigh)	1M	53.2
Claude Fable 5.1 (high)	1M	51.2
Claude Opus 5 (medium)	1M	44.8
Claude Sonnet 4.6 (max)	1M	30.1
Claude Fable 5.1 (medium)	1M	48.9
Claude Fable 5.1 (low)	1M	46.8
Claude Opus 4.6 (max)	1M	31.9
Claude Opus 4.7 (Non-reasoning)	1M	30.9
Claude Opus 5 (low)	1M	39.4
Claude Opus 4.5 (Reasoning)	200K	29.1
Claude Sonnet 4.6 (Non-reasoning)	1M	24.7
Claude Opus 4.6 (Non-reasoning)	1M	26.4
Claude Sonnet 4.6 (Non-reasoning)	1M	23.3
Claude Opus 4.5 (Non-reasoning)	200K	23.7
Claude 4.5 Haiku (Non-reasoning)	200K	15.4
Claude 4.5 Sonnet (Non-reasoning)	1M	19.3
Claude 4.1 Opus (Reasoning)	—	22.8
Claude 4 Opus (Reasoning)	—	20.6
Claude 4.1 Opus (Non-reasoning)	200K	18.6
Claude 3.7 Sonnet (Non-reasoning)	200K	15.3
Claude 4 Opus (Non-reasoning)	1M	16.6
Claude 4.5 Sonnet (Reasoning)	—	20.7
Claude 4.5 Haiku (Reasoning)	—	16.9
Claude 3.5 Sonnet (Oct '24)	200K	7.9
Claude 3 Haiku	200K	5.6
Claude 3 Opus	—	8.7
Claude 3.5 Sonnet (June '24)	—	7.2
Claude 2.0	—	5.5
Claude 2.1	—	5.6
Claude 3 Sonnet	200K	5.9
Claude 3.5 Haiku	200K	8.9
Claude 3.7 Sonnet (Reasoning)	200K	17.7
Claude 4 Sonnet (Non-reasoning)	1M	16.6
Claude 4 Sonnet (Reasoning)	—	18.9
Claude Instant	—	5.0