价格变动时间线
各家大模型 API 的调价 / 降价记录,按日期倒序 · 数据更新于 2026-10-22
📊 全部数据 · 总览每日自动比对各模型价格,有变动即记录在此。绿色为降价、红色为涨价; 点击模型名查看该模型完整价格历史曲线。← 返回总表
2026-09-17
- DeepSeek: DeepSeek V3最大输出16K → 16K
- DeepSeek: DeepSeek V3输出价$1.03 → $0.89-13%
- DeepSeek: DeepSeek V3输入价$0.26 → $0.32+24%
- NVIDIA: Nemotron 3 Nano 30B A3B输出价$0.20 → $0.24+20%
- NVIDIA: Nemotron 3 Nano 30B A3B输入价$0.05 → $0.06+20%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.01-10%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.14-10%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.07-10%
- NVIDIA: Nemotron 3 Ultra最大输出183K → 33K
- NVIDIA: Nemotron 3 Ultra缓存读$0.12 → $0.19+56%
- NVIDIA: Nemotron 3 Ultra输出价$2.40 → $3.13+30%
- NVIDIA: Nemotron 3 Ultra输入价$0.60 → $0.63+4%
- Meta: Muse Glimmer 30B输出价$1.50 → $1.10-27%
- Meta: Muse Glimmer 30B输入价$0.35 → $0.30-14%
- Z.ai: GLM 5.3 Flash缓存读$0.01 → $0.02+29%
- Z.ai: GLM 5.3 Flash输出价$0.23 → $0.30+29%
- Z.ai: GLM 5.3 Flash输入价$0.07 → $0.09+29%
2026-09-16
- qwen/qwen3.8-max缓存写$2.50 → —
- qwen/qwen3.8-max缓存读$0.25 → —
- qwen/qwen3.8-max输出价$6.00 → $0.00-100%
- qwen/qwen3.8-max输入价$2.00 → $0.00-100%
- anthropic/claude-opus-5-fast缓存写$12.50 → —
- anthropic/claude-opus-5-fast缓存读$1.00 → —
- MythoMax 13B输出价$0.06 → $0.11+83%
- MythoMax 13B输入价$0.06 → $0.08+33%
- MiniMax: MiniMax M1输入价$0.55 → $0.40-27%
- Mistral: Mistral Small 3.2 24B输出价$0.20 → $0.25+25%
- Mistral: Mistral Small 3.2 24B输入价$0.07 → $0.09+25%
- Qwen: Qwen3 VL 30B A3B Instruct最大输出16K → 33K
- Qwen: Qwen3 VL 30B A3B Instruct输出价$0.60 → $0.52-13%
- Qwen: Qwen3 VL 30B A3B Instruct输入价$0.15 → $0.13-13%
- Qwen: Qwen3.5-35B-A3B缓存读$0.16 → —
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02+2%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.15 → $0.16+2%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.08+2%
- MoonshotAI: Kimi K2.7 Code缓存读$0.15 → $0.18+20%
- MoonshotAI: Kimi K2.7 Code输出价$3.50 → $3.21-8%
- MoonshotAI: Kimi K2.7 Code输入价$0.71 → $0.71-1%
- Z.ai: GLM 5.2最大输出128K → 131K
- MoonshotAI: Kimi K3缓存读$0.30 → $0.30-1%
- MoonshotAI: Kimi K3输出价$13.28 → $15.00+13%
- MoonshotAI: Kimi K3输入价$2.65 → $3.00+13%
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.01 → $0.01+9%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.11 → $0.12+9%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.06 → $0.06+9%
- Z.ai: GLM 5.3 Flash缓存读$0.01 → $0.01-7%
- Z.ai: GLM 5.3 Flash输出价$0.25 → $0.23-7%
- Z.ai: GLM 5.3 Flash输入价$0.07 → $0.07-7%
2026-09-15
- qwen/qwen3.8-max最大输出131K → —
- qwen/qwen3.8-max上下文窗口1M → —
- anthropic/claude-opus-5-fast最大输出128K → —
- anthropic/claude-opus-5-fast上下文窗口1M → —
- Qwen: Qwen3 Next 80B A3B Thinking最大输出236K → 33K
- Qwen: Qwen3.5-35B-A3B最大输出16K → 66K
- Qwen: Qwen3.5-35B-A3B输出价$1.25 → $1.30+4%
- Qwen: Qwen3.5-35B-A3B输入价$0.31 → $0.16-48%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.01 → $0.02+2%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.15 → $0.15+2%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.07 → $0.08+2%
- Z.ai: GLM 5.2最大输出131K → 128K
- Z.ai: GLM 5.2缓存读$0.13 → $0.14+10%
- Z.ai: GLM 5.2输出价$2.15 → $4.40+105%
- Z.ai: GLM 5.2输入价$0.68 → $1.40+105%
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.01 → $0.01-8%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.12 → $0.11-8%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.06 → $0.06-8%
- Z.ai: GLM 5.3 Flash缓存读$0.03 → $0.01-50%
- Z.ai: GLM 5.3 Flash输出价$0.50 → $0.25-50%
- Z.ai: GLM 5.3 Flash输入价$0.15 → $0.07-50%
2026-09-14
- Meta: Llama 4 Maverick最大输出115K → 16K
- Meta: Llama 4 Maverick输出价$0.70 → $0.65-6%
- Meta: Llama 4 Maverick输入价$0.20 → $0.19-6%
- Qwen: Qwen3 14B最大输出8K → 16K
- Qwen: Qwen3 14B输出价$0.91 → $0.24-74%
- Qwen: Qwen3 14B输入价$0.23 → $0.12-47%
- NVIDIA: Nemotron 3 Super最大输出16K → 236K
- NVIDIA: Nemotron 3 Super输出价$0.40 → $0.45+12%
- NVIDIA: Nemotron 3 Super输入价$0.09 → $0.08-6%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.01 → $0.01+63%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.09 → $0.15+62%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.05 → $0.07+62%
- NVIDIA: Nemotron 3 Ultra最大输出33K → 183K
- NVIDIA: Nemotron 3 Ultra缓存读$0.19 → $0.12-36%
- NVIDIA: Nemotron 3 Ultra输出价$3.13 → $2.40-23%
- NVIDIA: Nemotron 3 Ultra输入价$0.63 → $0.60-4%
- Z.ai: GLM 5.2最大输出182K → 131K
- Z.ai: GLM 5.2缓存读$0.15 → $0.13-15%
- Z.ai: GLM 5.2输出价$2.00 → $2.15+7%
- Z.ai: GLM 5.2输入价$0.60 → $0.68+14%
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.01 → $0.01+50%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.08 → $0.12+50%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.04 → $0.06+50%
- DeepSeek: DeepSeek V4 Pro 0813最大输出384K → 393K
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.02 → $0.02-5%
- Z.ai: GLM 5.3最大输出131K → 944K
- Z.ai: GLM 5.3缓存读$0.20 → $0.26+28%
- Z.ai: GLM 5.3输出价$3.43 → $4.40+28%
- Z.ai: GLM 5.3输入价$1.09 → $1.40+28%
2026-09-13
- MoonshotAI: Kimi K2 0711最大输出100K → 98K
- Qwen: Qwen3 30B A3B Instruct 2507最大输出236K → 32K
- Qwen: Qwen3 30B A3B Instruct 2507输出价$0.30 → $0.19-36%
- Qwen: Qwen3 30B A3B Instruct 2507输入价$0.09 → $0.05-47%
- MoonshotAI: Kimi K2 0905最大输出100K → 98K
- MoonshotAI: Kimi K2 Thinking最大输出100K → 98K
- Google: Gemma 4 26B A4B 最大输出33K → 236K
- Google: Gemma 4 26B A4B 缓存读— → $0.05
- Google: Gemma 4 26B A4B 输出价$0.22 → $0.30+36%
- Google: Gemma 4 26B A4B 输入价$0.04 → $0.09+114%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.01 → $0.01-30%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.13 → $0.09-30%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.07 → $0.05-30%
- Meta: Muse Glimmer 30B输出价$1.10 → $1.50+36%
- Meta: Muse Glimmer 30B输入价$0.30 → $0.35+17%
- Z.ai: GLM 5.3最大输出944K → 131K
- Z.ai: GLM 5.3缓存读$0.26 → $0.20-22%
- Z.ai: GLM 5.3输出价$4.40 → $3.43-22%
- Z.ai: GLM 5.3输入价$1.40 → $1.09-22%
- Inference.net: Schematron V2 Small缓存读— → $0.05
- Inference.net: Schematron V2 Turbo缓存读— → $0.03
2026-09-12
- DeepSeek: DeepSeek V3 0324缓存读$0.11 → —
- Qwen: Qwen3 235B A22B Instruct 2507最大输出16K → 236K
- Qwen: Qwen3 235B A22B Instruct 2507缓存读— → $0.02
- Qwen: Qwen3 235B A22B Instruct 2507输出价$0.88 → $0.35-60%
- Qwen: Qwen3 235B A22B Instruct 2507输入价$0.22 → $0.09-60%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.01 → $0.01-3%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.14 → $0.13-3%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.07 → $0.07-3%
- DeepSeek: DeepSeek V4 Pro 0423最大输出384K → 393K
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.07 → $0.14+91%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.70 → $3.20+88%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.85 → $1.60+88%
- MoonshotAI: Kimi K3缓存读$0.21 → $0.30+47%
- MoonshotAI: Kimi K3输出价$9.01 → $13.28+47%
- MoonshotAI: Kimi K3输入价$1.80 → $2.65+47%
- Thinking Machines: Inkling最大输出33K → 472K
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.02 → $0.01-50%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.18 → $0.08-56%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.07 → $0.04-38%
- Qwen: Qwen3.8 27B缓存读$0.09 → $0.15+76%
- Qwen: Qwen3.8 27B输出价$3.00 → $2.55-15%
- Qwen: Qwen3.8 27B输入价$0.42 → $0.21-49%
2026-09-11
- DeepSeek: DeepSeek V3 0324输出价$1.14 → $1.00-12%
- DeepSeek: DeepSeek V3 0324输入价$0.29 → $0.25-14%
- Mistral: Mistral Small 3.2 24B上下文窗口131K → 256K
- Qwen: Qwen3 30B A3B Instruct 2507最大输出32K → 236K
- Qwen: Qwen3 30B A3B Instruct 2507输出价$0.19 → $0.30+55%
- Qwen: Qwen3 30B A3B Instruct 2507输入价$0.05 → $0.09+87%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.01-18%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.14-18%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.07-18%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.07 → $0.07-2%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.74 → $1.70-2%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.87 → $0.85-2%
- Z.ai: GLM 5.2最大输出128K → 182K
- Z.ai: GLM 5.2缓存读$0.05 → $0.15+188%
- Z.ai: GLM 5.2输出价$0.88 → $2.00+127%
- Z.ai: GLM 5.2输入价$0.28 → $0.60+114%
- MoonshotAI: Kimi K3缓存读$0.30 → $0.21-32%
- MoonshotAI: Kimi K3输出价$15.00 → $9.01-40%
- MoonshotAI: Kimi K3输入价$3.00 → $1.80-40%
- Upstage: Solar Pro 4缓存读$0.01 → $0.02+200%
- Upstage: Solar Pro 4输出价$0.12 → $0.36+200%
- Upstage: Solar Pro 4输入价$0.03 → $0.09+200%
- DeepSeek: DeepSeek V4 Pro 0813最大输出393K → 384K
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.02 → $0.02+5%
2026-09-10
- Qwen: Qwen3 30B A3B Instruct 2507最大输出236K → 32K
- Qwen: Qwen3 30B A3B Instruct 2507输出价$0.30 → $0.19-36%
- Qwen: Qwen3 30B A3B Instruct 2507输入价$0.09 → $0.05-47%
- Z.ai: GLM 4.7 Flash上下文窗口203K → 200K
- MiniMax: MiniMax M2.5最大输出131K → 128K
- MiniMax: MiniMax M2.5缓存读$0.03 → $0.03-10%
- MiniMax: MiniMax M2.5输出价$1.20 → $1.08-10%
- MiniMax: MiniMax M2.5输入价$0.30 → $0.27-10%
- Google: Gemma 4 26B A4B 最大输出16K → 33K
- Google: Gemma 4 26B A4B 输出价$0.34 → $0.22-35%
- Google: Gemma 4 26B A4B 输入价$0.07 → $0.04-40%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02-7%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.18 → $0.16-7%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.09 → $0.08-7%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.08 → $0.07-9%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.91 → $1.74-9%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.96 → $0.87-9%
- Z.ai: GLM 5.2最大输出131K → 128K
- Z.ai: GLM 5.2缓存读$0.19 → $0.05-73%
- Z.ai: GLM 5.2输出价$3.04 → $0.88-71%
- Z.ai: GLM 5.2输入价$0.97 → $0.28-71%
- Tencent: Hy3缓存读$0.03 → $0.02-38%
- Tencent: Hy3输出价$0.53 → $0.33-37%
- Tencent: Hy3输入价$0.13 → $0.08-37%
- DeepSeek: DeepSeek V4 Pro 0813最大输出384K → 393K
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.04 → $0.02-47%
- DeepSeek: DeepSeek V4 Pro 0813输出价$3.15 → $1.74-45%
- DeepSeek: DeepSeek V4 Pro 0813输入价$1.05 → $0.58-45%
- Z.ai: GLM 5.3 Flash缓存读$0.01 → $0.03+100%
- Z.ai: GLM 5.3 Flash输出价$0.25 → $0.50+100%
- Z.ai: GLM 5.3 Flash输入价$0.07 → $0.15+100%
- DeepSeek: DeepSeek V4.1 Flash缓存读$0.01 → $0.00-50%
- DeepSeek: DeepSeek V4.1 Flash输出价$1.20 → $0.60-50%
- DeepSeek: DeepSeek V4.1 Flash输入价$0.30 → $0.15-50%
- DeepSeek: DeepSeek V3最大输出16K → 16K
- DeepSeek: DeepSeek V3输出价$0.89 → $1.03+16%
- DeepSeek: DeepSeek V3输入价$0.32 → $0.26-20%
- Qwen: Qwen3 30B A3B Instruct 2507最大输出32K → 236K
- Qwen: Qwen3 30B A3B Instruct 2507输出价$0.19 → $0.30+55%
- Qwen: Qwen3 30B A3B Instruct 2507输入价$0.05 → $0.09+87%
- Qwen: Qwen3 Next 80B A3B Instruct缓存读$0.07 → —
- MoonshotAI: Kimi K2 Thinking最大输出236K → 100K
- MiniMax: MiniMax M2.5最大输出128K → 131K
- MiniMax: MiniMax M2.5缓存读$0.03 → $0.03+11%
- MiniMax: MiniMax M2.5输出价$1.08 → $1.20+11%
- MiniMax: MiniMax M2.5输入价$0.27 → $0.30+11%
- Qwen: Qwen3.5-122B-A10B最大输出82K → 66K
- Qwen: Qwen3.5-122B-A10B输出价$2.40 → $2.08-13%
- Qwen: Qwen3.5-122B-A10B输入价$0.29 → $0.26-10%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.01 → $0.02+26%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.14 → $0.18+27%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.07 → $0.09+27%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.07 → $0.08+10%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.74 → $1.91+10%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.87 → $0.96+10%
- Z.ai: GLM 5.2最大输出128K → 131K
- Z.ai: GLM 5.2缓存读$0.05 → $0.19+272%
- Z.ai: GLM 5.2输出价$0.88 → $3.04+245%
- Z.ai: GLM 5.2输入价$0.28 → $0.97+245%
- Tencent: Hy3缓存读$0.02 → $0.03+60%
- Tencent: Hy3输出价$0.33 → $0.53+60%
- Tencent: Hy3输入价$0.08 → $0.13+60%
- Thinking Machines: Inkling最大输出472K → 33K
- DeepSeek: DeepSeek V4 Pro 0813最大输出393K → 384K
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.02 → $0.04+90%
- DeepSeek: DeepSeek V4 Pro 0813输出价$1.74 → $3.15+81%
- DeepSeek: DeepSeek V4 Pro 0813输入价$0.58 → $1.05+81%
- Z.ai: GLM 5.3 Flash缓存读$0.01 → $0.01+7%
- Z.ai: GLM 5.3 Flash输出价$0.23 → $0.25+7%
- Z.ai: GLM 5.3 Flash输入价$0.07 → $0.07+7%
2026-09-09
- MiniMax: MiniMax M1输入价$0.40 → $0.55+38%
- Qwen: Qwen3 235B A22B Instruct 2507输出价$0.55 → $0.88+60%
- Qwen: Qwen3 235B A22B Instruct 2507输入价$0.09 → $0.22+144%
- Qwen: Qwen3 Next 80B A3B Instruct最大输出236K → 16K
- Qwen: Qwen3 Next 80B A3B Instruct输入价$0.10 → $0.09-10%
- Z.ai: GLM 4.6最大输出131K → 16K
- Z.ai: GLM 4.6缓存读$0.11 → $0.08-27%
- Z.ai: GLM 4.6输出价$2.20 → $1.75-20%
- Z.ai: GLM 4.6输入价$0.55 → $0.43-22%
- MoonshotAI: Kimi K2 Thinking最大输出100K → 236K
- Z.ai: GLM 4.7 Flash缓存读$0.01 → —
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.01-15%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.14-15%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.07-15%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.08 → $0.07-8%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.89 → $1.74-8%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.94 → $0.87-8%
- Z.ai: GLM 5.2最大输出131K → 128K
- Z.ai: GLM 5.2缓存读$0.19 → $0.05-73%
- Z.ai: GLM 5.2输出价$3.04 → $0.88-71%
- Z.ai: GLM 5.2输入价$0.97 → $0.28-71%
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.06 → $0.02-68%
- Qwen: Qwen3.8 2.4T A95B最大输出262K → 131K
- DeepSeek: DeepSeek V4 Flash Vision Exp最大输出384K → 944K
- Z.ai: GLM 5.3 Flash缓存读$0.01 → $0.01-7%
- Z.ai: GLM 5.3 Flash输出价$0.25 → $0.23-7%
- Z.ai: GLM 5.3 Flash输入价$0.07 → $0.07-7%
- IBM: Granite 4.2 8B缓存读$0.05 → $0.01-70%
- IBM: Granite 4.2 8B输出价$0.15 → $0.25+67%
- IBM: Granite 4.2 8B输入价$0.10 → $0.06-40%
2026-09-08
- DeepSeek: DeepSeek V3 0324缓存读— → $0.11
- DeepSeek: DeepSeek V3 0324输出价$1.00 → $1.14+14%
- DeepSeek: DeepSeek V3 0324输入价$0.25 → $0.29+16%
- Qwen: Qwen3 14B最大输出16K → 8K
- Qwen: Qwen3 14B输出价$0.24 → $0.91+279%
- Qwen: Qwen3 14B输入价$0.12 → $0.23+90%
- MiniMax: MiniMax M1输入价$0.55 → $0.40-27%
- DeepSeek: DeepSeek V3.1最大输出145K → 33K
- DeepSeek: DeepSeek V3.1缓存读$0.55 → $0.13-76%
- DeepSeek: DeepSeek V3.1输出价$1.65 → $0.95-42%
- DeepSeek: DeepSeek V3.1输入价$0.55 → $0.25-55%
- Z.ai: GLM 4.6最大输出16K → 131K
- Z.ai: GLM 4.6缓存读$0.08 → $0.11+38%
- Z.ai: GLM 4.6输出价$1.75 → $2.20+26%
- Z.ai: GLM 4.6输入价$0.43 → $0.55+28%
- Z.ai: GLM 4.7 Flash最大输出16K → 118K
- Z.ai: GLM 4.7 Flash输入价$0.06 → $0.06+1%
- Qwen: Qwen3.5 397B A17B最大输出66K → 236K
- Qwen: Qwen3.5 397B A17B缓存读— → $0.23
- Qwen: Qwen3.5 397B A17B输出价$2.34 → $3.50+50%
- Qwen: Qwen3.5 397B A17B输入价$0.39 → $0.55+41%
- NVIDIA: Nemotron 3 Super上下文窗口1M → 262K
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02-7%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.18 → $0.16-7%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.09 → $0.08-7%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.08 → $0.08-1%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.91 → $1.89-1%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.96 → $0.94-1%
- MoonshotAI: Kimi K2.7 Code缓存读$0.18 → $0.15-17%
- MoonshotAI: Kimi K2.7 Code输出价$3.40 → $3.50+3%
- MoonshotAI: Kimi K2.7 Code输入价$0.66 → $0.71+8%
- DeepSeek: DeepSeek V4 Flash 0731最大输出131K → 944K
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.03 → $0.02-43%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.28 → $0.18-36%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.14 → $0.07-54%
- DeepSeek: DeepSeek V4 Pro 0813最大输出384K → 393K
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.02 → $0.06+163%
- DeepSeek: DeepSeek V4 Pro 0813输出价$1.98 → $1.74-12%
- DeepSeek: DeepSeek V4 Pro 0813输入价$0.66 → $0.58-12%
2026-09-07
- Qwen: Qwen3 Next 80B A3B Thinking最大输出33K → 236K
- Z.ai: GLM 4.6最大输出131K → 16K
- Z.ai: GLM 4.6缓存读$0.11 → $0.08-27%
- Z.ai: GLM 4.6输出价$2.20 → $1.75-20%
- Z.ai: GLM 4.6输入价$0.55 → $0.43-22%
- Qwen: Qwen3.5 397B A17B缓存读$0.23 → —
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02+12%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.18+12%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.09+12%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.05 → $0.08+55%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.23 → $1.91+55%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.62 → $0.96+55%
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.01 → $0.03+180%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.10 → $0.28+180%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.05 → $0.14+180%
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.02 → $0.02+14%
- DeepSeek: DeepSeek V4 Pro 0813输出价$1.74 → $1.98+14%
- DeepSeek: DeepSeek V4 Pro 0813输入价$0.58 → $0.66+14%
- Z.ai: GLM 5.3最大输出262K → 944K
- Z.ai: GLM 5.3缓存读$0.14 → $0.26+86%
2026-09-06
- TheDrummer: UnslopNemo 12B最大输出26K → 819K
- Qwen: Qwen3 235B A22B Instruct 2507缓存读$0.02 → —
- Qwen: Qwen3.5 397B A17B最大输出236K → 66K
- Qwen: Qwen3.5 397B A17B输出价$3.50 → $2.34-33%
- Qwen: Qwen3.5 397B A17B输入价$0.55 → $0.39-29%
- Qwen: Qwen3.5-35B-A3B缓存读$0.25 → $0.16-37%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02-5%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.17 → $0.16-5%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.08-5%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.07 → $0.05-24%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.62 → $1.23-24%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.81 → $0.62-24%
- DeepSeek: DeepSeek V4 Flash 0731最大输出944K → 131K
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.02 → $0.01-37%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.18 → $0.10-44%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.07 → $0.05-23%
2026-09-05
- Qwen: Qwen3 235B A22B Instruct 2507最大输出236K → 16K
- Qwen: Qwen3 235B A22B Instruct 2507输出价$0.35 → $0.55+57%
- Qwen: Qwen3 235B A22B Instruct 2507输入价$0.09 → $0.09+3%
- Qwen: Qwen3.5-35B-A3B输出价$0.75 → $1.25+67%
- Qwen: Qwen3.5-35B-A3B输入价$0.08 → $0.31+291%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02-5%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.17 → $0.17-5%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.09 → $0.08-5%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.08 → $0.07-17%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.94 → $1.62-17%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.97 → $0.81-17%
2026-09-04
- ReMM SLERP 13B最大输出4K → 6K
- ReMM SLERP 13B输入价$0.45 → $0.35-22%
- Meta: Llama 3.3 70B Instruct缓存读$0.71 → —
- Meta: Llama 3.3 70B Instruct输入价$0.71 → $0.10-86%
- Qwen: Qwen3.5-35B-A3B最大输出236K → 16K
- Qwen: Qwen3.5-35B-A3B输出价$1.25 → $0.75-40%
- Qwen: Qwen3.5-35B-A3B输入价$0.25 → $0.08-68%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02+10%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.17+10%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.09+10%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.09 → $0.08-5%
- DeepSeek: DeepSeek V4 Pro 0423输出价$2.05 → $1.94-5%
- DeepSeek: DeepSeek V4 Pro 0423输入价$1.02 → $0.97-5%
- Qwen: Qwen3.6 27B最大输出236K → 66K
- Qwen: Qwen3.6 27B缓存读$0.12 → $0.03-75%
- Qwen: Qwen3.6 27B输出价$3.60 → $2.00-44%
- Qwen: Qwen3.6 27B输入价$0.60 → $0.30-50%
- NVIDIA: Nemotron 3 Ultra最大输出183K → 33K
- NVIDIA: Nemotron 3 Ultra缓存读$0.12 → $0.19+56%
- NVIDIA: Nemotron 3 Ultra输出价$2.40 → $3.13+30%
- NVIDIA: Nemotron 3 Ultra输入价$0.60 → $0.63+4%
- DeepSeek: DeepSeek V4 Pro 0813缓存读$0.02 → $0.02-12%
- DeepSeek: DeepSeek V4 Pro 0813输出价$1.98 → $1.74-12%
- DeepSeek: DeepSeek V4 Pro 0813输入价$0.66 → $0.58-12%
- Qwen: Qwen3.8 27B缓存写$0.53 → —
2026-09-03
- Meta: Llama 3.3 70B Instruct最大输出115K → 16K
- Meta: Llama 3.3 70B Instruct输出价$0.71 → $0.32-55%
- DeepSeek: DeepSeek V3最大输出16K → 16K
- DeepSeek: DeepSeek V3输出价$1.03 → $0.89-13%
- DeepSeek: DeepSeek V3输入价$0.26 → $0.32+24%
- Qwen: Qwen2.5 VL 72B Instruct最大输出29K → 115K
- Qwen: Qwen2.5 VL 72B Instruct缓存读— → $0.40
- Qwen: Qwen2.5 VL 72B Instruct输出价$0.75 → $1.00+33%
- Qwen: Qwen2.5 VL 72B Instruct输入价$0.25 → $0.80+220%
- DeepSeek: DeepSeek V3.1最大输出33K → 145K
- DeepSeek: DeepSeek V3.1缓存读$0.13 → $0.55+323%
- DeepSeek: DeepSeek V3.1输出价$0.95 → $1.65+74%
- DeepSeek: DeepSeek V3.1输入价$0.25 → $0.55+120%
- Z.ai: GLM 4.6最大输出16K → 131K
- Z.ai: GLM 4.6缓存读$0.08 → $0.11+38%
- Z.ai: GLM 4.6输出价$1.75 → $2.20+26%
- Z.ai: GLM 4.6输入价$0.43 → $0.55+28%
- NVIDIA: Nemotron 3 Nano 30B A3B最大输出228K → 236K
- NVIDIA: Nemotron 3 Nano 30B A3B缓存读$0.03 → $0.03+20%
- Qwen: Qwen3.5 397B A17B最大输出66K → 236K
- Qwen: Qwen3.5 397B A17B缓存读— → $0.23
- Qwen: Qwen3.5 397B A17B输出价$2.34 → $3.50+50%
- Qwen: Qwen3.5 397B A17B输入价$0.39 → $0.55+41%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.07 → $0.09+18%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.74 → $2.05+18%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.87 → $1.02+18%
- Meta: Muse Glimmer 30B最大输出16K → 118K
- Meta: Muse Glimmer 30B输出价$1.20 → $1.10-8%
- Qwen: Qwen3.8 2.4T A95B最大输出131K → 262K
- Qwen: Qwen3.8 2.4T A95B缓存读$0.20 → $0.25+25%
- Qwen: Qwen3.8 27B输出价$2.55 → $3.00+18%
- Qwen: Qwen3.8 27B输入价$0.42 → $0.42-1%
- Z.ai: GLM 5.3最大输出131K → 262K
- Z.ai: GLM 5.3缓存读$0.26 → $0.14-46%
2026-09-02
- Meta: Llama 4 Scout缓存读$0.06 → —
- DeepSeek: DeepSeek V3.1最大输出145K → 33K
- DeepSeek: DeepSeek V3.1缓存读$0.55 → $0.13-76%
- DeepSeek: DeepSeek V3.1输出价$1.65 → $0.95-42%
- DeepSeek: DeepSeek V3.1输入价$0.55 → $0.25-55%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.01 → $0.02+11%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.14 → $0.16+11%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.07 → $0.08+11%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.09 → $0.07-15%
- DeepSeek: DeepSeek V4 Pro 0423输出价$2.05 → $1.74-15%
- DeepSeek: DeepSeek V4 Pro 0423输入价$1.03 → $0.87-15%
- NVIDIA: Nemotron 3 Ultra最大输出16K → 183K
- NVIDIA: Nemotron 3 Ultra缓存读$0.10 → $0.12+20%
- NVIDIA: Nemotron 3 Ultra输出价$2.20 → $2.40+9%
- NVIDIA: Nemotron 3 Ultra输入价$0.50 → $0.60+20%
- Z.ai: GLM 5.2最大输出262K → 131K
- Z.ai: GLM 5.2缓存读$0.22 → $0.19-13%
- Z.ai: GLM 5.2输出价$3.74 → $3.04-19%
- Z.ai: GLM 5.2输入价$1.19 → $0.97-19%
- Qwen: Qwen3.8 2.4T A95B最大输出262K → 131K
- Qwen: Qwen3.8 2.4T A95B缓存读$0.25 → $0.20-20%
2026-09-01
- Mancer: Weaver (alpha)输入价$0.50 → $0.40-20%
- Magnum v4 72B输入价$3.00 → $2.50-17%
- Meta: Llama 4 Scout最大输出8K → 16K
- Meta: Llama 4 Scout输出价$0.34 → $0.30-12%
- Meta: Llama 4 Scout输入价$0.11 → $0.10-9%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.01-10%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.14-10%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.07-10%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.07 → $0.09+18%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.74 → $2.05+18%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.87 → $1.03+18%
2026-08-31
- Meta: Llama 4 Maverick最大输出16K → 115K
- Meta: Llama 4 Maverick输出价$0.80 → $0.70-13%
- Qwen: Qwen3 Next 80B A3B Instruct最大输出16K → 236K
- Qwen: Qwen3 Next 80B A3B Instruct输入价$0.09 → $0.10+11%
- MoonshotAI: Kimi K2.5缓存读$0.10 → $0.07-30%
- MoonshotAI: Kimi K2.5输出价$3.00 → $2.25-25%
- MoonshotAI: Kimi K2.5输入价$0.60 → $0.45-25%
- Kwaipilot: KAT-Coder-Pro V2最大输出80K → 144K
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02+1%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.16+1%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.08+1%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.03 → $0.07+108%
- DeepSeek: DeepSeek V4 Pro 0423输出价$0.83 → $1.74+109%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.42 → $0.87+108%
- Kwaipilot: KAT-Coder-Pro V2.5最大输出80K → 236K
2026-08-30
- OpenAI: GPT-4.1 Nano最大输出943K → 33K
- OpenAI: GPT-4.1 Nano缓存读$0.03 → $0.03-17%
- Qwen: Qwen3 Next 80B A3B Instruct最大输出236K → 16K
- Qwen: Qwen3 Next 80B A3B Instruct输入价$0.10 → $0.09-10%
- Mistral: Devstral 2 2512缓存读$0.04 → $0.04-9%
- Mistral: Devstral 2 2512输出价$2.20 → $2.00-9%
- Mistral: Devstral 2 2512输入价$0.44 → $0.40-9%
- Arcee AI: Trinity Large Thinking最大输出236K → 80K
- Arcee AI: Trinity Large Thinking输出价$0.85 → $0.80-6%
- Arcee AI: Trinity Large Thinking输入价$0.22 → $0.25+14%
- Z.ai: GLM 5.1最大输出182K → 128K
- Z.ai: GLM 5.1缓存读$0.23 → $0.18-23%
- Z.ai: GLM 5.1输出价$3.96 → $3.04-23%
- Z.ai: GLM 5.1输入价$1.26 → $0.97-23%
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02-4%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.16 → $0.16-4%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.08 → $0.08-4%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.04 → $0.03-22%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.08 → $0.83-23%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.54 → $0.42-23%
- Thinking Machines: Inkling最大输出262K → 472K
- Thinking Machines: Inkling缓存读$0.16 → $0.17+6%
- Thinking Machines: Inkling输入价$0.95 → $1.00+5%
2026-08-29
- DeepSeek: DeepSeek V4 Flash 0423缓存读$0.02 → $0.02-5%
- DeepSeek: DeepSeek V4 Flash 0423输出价$0.17 → $0.16-5%
- DeepSeek: DeepSeek V4 Flash 0423输入价$0.09 → $0.08-5%
- DeepSeek: DeepSeek V4 Pro 0423缓存读$0.06 → $0.04-26%
- DeepSeek: DeepSeek V4 Pro 0423输出价$1.46 → $1.08-26%
- DeepSeek: DeepSeek V4 Pro 0423输入价$0.73 → $0.54-26%
- DeepSeek: DeepSeek V4 Flash 0731缓存读$0.01 → $0.02+60%
- DeepSeek: DeepSeek V4 Flash 0731输出价$0.10 → $0.18+80%
- DeepSeek: DeepSeek V4 Flash 0731输入价$0.05 → $0.07+30%
2026-08-28
- Google: Gemma 3 27B上下文窗口262K → 131K
- OpenAI: GPT-4.1 Nano最大输出33K → 943K
- OpenAI: GPT-4.1 Nano缓存读$0.03 → $0.03+20%
- Qwen: Qwen3 VL 30B A3B Instruct最大输出33K → 16K
- Qwen: Qwen3 VL 30B A3B Instruct输出价$0.52 → $0.60+15%
- Qwen: Qwen3 VL 30B A3B Instruct输入价$0.13 → $0.15+15%
- Mistral: Voxtral Small 24B 2507最大输出26K → 26K
- Mistral: Voxtral Small 24B 2507上下文窗口32K → 33K
- DeepSeek: DeepSeek V3.2最大输出147K → 66K
- DeepSeek: DeepSeek V3.2缓存读$0.13 → $0.13+3%
- DeepSeek: DeepSeek V3.2输出价$0.38 → $0.40+5%
价格数据参考 OpenRouter 及各厂商公开页面整理,以 美元 / 百万 tokens 计,可能滞后或有误,请以官方页面为准。不构成任何采购或投资建议。