4,854 个接口 · 11,193 个价格 · 最近采集 23 分钟前 · 30 天内 442 条需关注的调价
每个平台公布的每一个价格
采集器定时抓取 10 个平台公开的全部价格（默认每日一次），并完整保存原始页面快照，绝无任何手工篡改。搜索具体模型，即可查看当前各单位下的最低价格。
模型
Claude 
Gemini 
Qwen 
DeepSeek 
GLM 
Kimi 
Mistral 
MiniMax 
Hunyuan 
Nemotron 
Llama 第 1–100 条，共 594 条
每百万缓存读取 token100
| 平台 | 模型 · 计价项 | 美元 | 平台原文 | 生效于 |
|---|---|---|---|---|
| OpenRouter第三方 | DeepSeek: DeepSeek V4.1 Flash 1017 (batch) | $0.002/百万缓存读最低 | input_cache_read: 0.000000002 / token (endpoint morph/fp8) | |
| OpenRouter官方 | Meta: Muse Spark 1.2 Contributor | $0.002/百万缓存读最低 | input_cache_read: 0.000000002 / token (endpoint meta) | |
| OpenRouter官方 | Meta: Muse Spark 1.3 Contributor | $0.002/百万缓存读最低 | input_cache_read: 0.000000002 / token (endpoint meta) | |
| Aiberm未披露 | DeepSeek V4.1 Flash | $0.0024/百万缓存读 | model_ratio 0.26087 × $2/M · group default ×0.23 · cache_ratio 0.02 | |
| KIE未披露 | DeepSeek V4.1 Flash | $0.0025/百万缓存读 | 0.5 credits per million tokens ($0.0025) | |
| OpenRouter官方 | Nex AGI: Nex-N2.5-Mini | $0.0025/百万缓存读 | input_cache_read: 0.0000000025 / token | |
| OpenRouter官方 | OpenAI: GPT-5 Nano (batch) | $0.0025/百万缓存读 | input_cache_read: 0.0000000025 / token (endpoint openai) | |
| OpenRouter第三方 | Xiaomi: MiMo-V2.6-Flashvia Darkbloom · fp4 | $0.0028/百万缓存读 | input_cache_read: 0.0000000028 / token (endpoint darkbloom/fp4) | |
| KIE未披露 | gpt-6-luna | $0.003/百万缓存读 | 0.6 credits per million tokens ($0.003)平台声称官方或 Fal 价 $0.01/百万缓存读 | |
| OpenRouter官方 | inclusionAI: Ling 3.0 Tiny | $0.003/百万缓存读 | input_cache_read: 0.000000003 / token (endpoint antdigital/fp8) | |
| OpenRouter官方 | Particle AI: Particle-1 | $0.003/百万缓存读 | input_cache_read: 0.000000003 / token | |
| OpenRouter第三方 | Xiaomi: MiMo-V2.6-Provia GMICloud · bf16 | $0.003/百万缓存读 | input_cache_read: 0.000000003 / token (endpoint gmicloud/bf16) | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4.1 Flash (batch) | $0.0034/百万缓存读 | input_cache_read: 0.00000000336 / token (endpoint deepinfra/fp8) | |
| OpenRouter第三方 | Xiaomi: MiMo-V2.5-Provia AtlasCloud · fp8 | $0.0036/百万缓存读 | input_cache_read: 0.0000000036 / token (endpoint atlas-cloud/fp8) | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4.1 Flashvia Morph · fp8 | $0.004/百万缓存读 | input_cache_read: 0.000000004 / token (endpoint morph/fp8) | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4.1 Flash 1017via Morph · fp8 | $0.004/百万缓存读 | input_cache_read: 0.000000004 / token (endpoint morph/fp8) | |
| OpenRouter第三方 | Meta: Muse Embervia Google · global | $0.004/百万缓存读 | input_cache_read: 0.000000004 / token (endpoint google-vertex/global) | |
| OpenRouter第三方 | inclusionAI: Ling 3.0 Flashvia Novita | $0.0042/百万缓存读 | input_cache_read: 0.0000000042 / token (endpoint novita) | |
| OpenRouter官方 | Anthropic: Claude Haiku 5.5 (batch) | $0.005/百万缓存读 | input_cache_read: 0.000000005 / token (endpoint anthropic) | |
| OpenRouter官方 | OpenAI: GPT-6 Luna Pro (batch) | $0.005/百万缓存读 | input_cache_read: 0.000000005 / token (endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-6 Luna (batch) | $0.005/百万缓存读 | input_cache_read: 0.000000005 / token (endpoint openai) | |
| OpenRouter官方 | Upstage: Solar Mini 4via Upstage · zdr | $0.005/百万缓存读 | input_cache_read: 0.000000005 / token (endpoint upstage/zdr) | |
| OpenRouter未披露 | DeepSeek: DeepSeek Flash Latest | $0.005/百万缓存读 | input_cache_read: 0.000000005 / token | |
| KIE未披露 | gpt-5.6-luna | $0.0056/百万缓存读 | 1.12 credits per million tokens ($0.0056)平台声称官方或 Fal 价 $0.02/百万缓存读 | |
| OpenRouter第三方 | Meituan: LongCat 2.0 | $0.006/百万缓存读 | input_cache_read: 0.000000006 / token (endpoint atlas-cloud/fp8) | |
| OpenRouter官方 | Qwen: Qwen3.7 Flashvia Alibaba | $0.006/百万缓存读 | input_cache_read: 0.000000006 / token (endpoint alibaba) | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4 Flash Vision Expvia DeepInfra · fp8 | $0.0069/百万缓存读 | input_cache_read: 0.00000000686 / token (endpoint deepinfra/fp8) | |
| Aiberm未披露 | GPT-4o Mini | $0.0069/百万缓存读 | model_ratio 0.05 × $2/M · group default ×0.23 · cache_ratio 0.3 | |
| OpenRouter第三方 | Qwen: Qwen3.8 27Bvia PrimeIntellect · fp8 | $0.007/百万缓存读 | input_cache_read: 0.000000007 / token (endpoint primeintellect/fp8) | |
| Aiberm未披露 | Claude Haiku 5.5 | $0.0074/百万缓存读 | model_ratio 0.16 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| OpenRouter第三方 | Google: Gemma 4 26B A4B via PrimeIntellect · bf16 | $0.008/百万缓存读 | input_cache_read: 0.000000008 / token (endpoint primeintellect/bf16) | |
| OpenRouter第三方 | inclusionAI: Ling 3.1 Flash | $0.009/百万缓存读 | input_cache_read: 0.000000009 / token (endpoint novita) | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4 Flash 0423via Baidu · fp8 | $0.01/百万缓存读 | input_cache_read: 0.000000009996 / token (endpoint baidu/fp8) | |
| OpenRouter官方合作 | Anthropic: Claude Haiku 5.5via Google · global | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint google-vertex/global) | |
| OpenRouter官方 | Google: Gemini 2.5 Flash Litevia Google AI Studio | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint google-ai-studio) | |
| OpenRouter官方 | Mistral: Ministral 3 3B 2512via Mistral · zdr | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint mistral/zdr) | |
| OpenRouter官方 | Mistral: Voxtral Small 24B 2507 | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token | |
| OpenRouter第三方 | NVIDIA: Nemotron 3.5 Super VLvia Darkbloom · nvfp4 | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint darkbloom/nvfp4) | |
| OpenRouter官方合作 | OpenAI: GPT-5 Nanovia Azure | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint azure) | |
| OpenRouter官方 | OpenAI: GPT-5.4 Nano (batch) | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-5.6 Luna Pro (batch) | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-5.6 Luna (batch) | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint openai) | |
| OpenRouter官方合作 | OpenAI: GPT-6 Lunavia Amazon Bedrock · global | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint amazon-bedrock/global) | |
| OpenRouter官方合作 | OpenAI: GPT-6 Luna Provia OpenAI | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-6 Luna Pro (batch)提示词超过 272K | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (>272000; endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-6 Luna (batch)提示词超过 272K | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token (>272000; endpoint openai) | |
| OpenRouter未披露 | Anthropic: Claude Haiku Latest | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token | |
| OpenRouter未披露 | OpenAI: GPT Luna Latest | $0.01/百万缓存读 | input_cache_read: 0.00000001 / token | |
| Runware未披露 | Claude Haiku 5.5≤ 100K context | $0.01/百万缓存读 | $1e-8 / cachedInputToken · ≤ 100K context | |
| Runware未披露 | DeepSeek-V4.1-FlashPer 1M | $0.01/百万缓存读 | $1e-8 / cachedInputToken | |
| Runware未披露 | DeepSeek-V4.1-Flash-1017Per 1M | $0.01/百万缓存读 | $1e-8 / cachedInputToken | |
| Runware未披露 | GPT-6 Luna≤ 272K context | $0.01/百万缓存读 | $1e-8 / cachedInputToken · ≤ 272K context | |
| OpenRouter第三方 | Mistral: Mistral Nemovia Io Net · fp16 | $0.011/百万缓存读 | input_cache_read: 0.0000000108 / token (endpoint io-net/fp16) | |
| OpenRouter第三方 | MiniMax: MiniMax M3.1 Flashvia GMICloud · fp8 | $0.011/百万缓存读 | input_cache_read: 0.00000001125 / token (endpoint gmicloud/fp8) | |
| Aiberm未披露 | Claude Haiku 4.5 | $0.012/百万缓存读 | model_ratio 0.25 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| Aiberm未披露 | Claude Haiku 4.5 20251001 | $0.012/百万缓存读 | model_ratio 0.25 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| Aiberm未披露 | Claude Haiku 4.5 20251001 Thinking | $0.012/百万缓存读 | model_ratio 0.25 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4 Flash 0731 (batch) | $0.012/百万缓存读 | input_cache_read: 0.000000012 / token (endpoint deepinfra/fp8) | |
| OpenRouter第三方 | inclusionAI: Ling 3.0 Flash VL | $0.012/百万缓存读 | input_cache_read: 0.000000012 / token (endpoint deepinfra/fp16) | |
| OpenRouter第三方 | NVIDIA: Nemotron 3 Nano 30B A3B | $0.012/百万缓存读 | input_cache_read: 0.000000012 / token (endpoint nebius/fp8) | |
| Runware未披露 | Gemma 4 31BPer 1M | $0.012/百万缓存读 | $1.2e-8 / cachedInputToken | |
| OpenRouter官方 | Google: Gemini 3.1 Flash Lite (batch) | $0.013/百万缓存读 | input_cache_read: 0.0000000125 / token (endpoint google-vertex/global) | |
| OpenRouter第三方 | NVIDIA: Nemotron 3.5 Lightningvia Darkbloom · int4 | $0.013/百万缓存读 | input_cache_read: 0.0000000125 / token (endpoint darkbloom/int4) | |
| OpenRouter官方 | OpenAI: GPT-4.1 Nano (batch) | $0.013/百万缓存读 | input_cache_read: 0.0000000125 / token (endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-5 Mini (batch) | $0.013/百万缓存读 | input_cache_read: 0.0000000125 / token (endpoint openai) | |
| OpenRouter第三方 | DeepSeek: DeepSeek V4 Flash 0731via Relace · fp4 | $0.014/百万缓存读 | input_cache_read: 0.000000014 / token (endpoint relace/fp4) | |
| Runware未披露 | DeepSeek-V4-Flash-0731Per 1M | $0.014/百万缓存读 | $1.4e-8 / cachedInputToken | |
| KIE未披露 | gpt-6-1-terra | $0.015/百万缓存读 | 3 credits per million tokens ($0.015)平台声称官方或 Fal 价 $0.05/百万缓存读 | |
| OpenRouter官方 | Google: Gemini 3.5 Flash Lite (batch) | $0.015/百万缓存读 | input_audio_cache: 0.000000015 / token (endpoint google-vertex/global) | |
| OpenRouter官方 | Google: Gemini 3.5 Flash Lite (batch) | $0.015/百万缓存读 | input_cache_read: 0.000000015 / token (endpoint google-vertex/global) | |
| OpenRouter第三方 | IBM: Granite 4.2 8Bvia CoreWeave · bf16 | $0.015/百万缓存读 | input_cache_read: 0.000000015 / token (endpoint coreweave/bf16) | |
| OpenRouter官方 | Mistral: Ministral 3 8B 2512via Mistral · zdr | $0.015/百万缓存读 | input_cache_read: 0.000000015 / token (endpoint mistral/zdr) | |
| OpenRouter官方 | Mistral: Mistral Small 4via Mistral · zdr | $0.015/百万缓存读 | input_cache_read: 0.000000015 / token (endpoint mistral/zdr) | |
| OpenRouter官方 | Nex AGI: Nex-N2.5-Pro | $0.015/百万缓存读 | input_cache_read: 0.000000015 / token (endpoint nex-agi/fp8) | |
| Runware未披露 | MiniMax M3.1 FlashPer 1M | $0.015/百万缓存读 | $1.5e-8 / cachedInputToken | |
| Aiberm未披露 | Claude Sonnet 5.5 | $0.015/百万缓存读 | model_ratio 0.6522 × $2/M · group default ×0.23 · cache_ratio 0.05 | |
| OpenRouter官方 | Qwen: Qwen3.8 Flashvia Alibaba | $0.016/百万缓存读 | input_cache_read: 0.000000016 / token (endpoint alibaba) | |
| OpenRouter官方 | Qwen: Qwen3.8 Omni Flash | $0.016/百万缓存读 | input_cache_read: 0.000000016 / token (endpoint alibaba) | |
| Aiberm未披露 | GPT-6.1 Terra | $0.016/百万缓存读 | model_ratio 0.35325 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| Aiberm未披露 | GPT-4.1 Mini | $0.018/百万缓存读 | model_ratio 0.13 × $2/M · group default ×0.23 · cache_ratio 0.3 | |
| OpenRouter第三方 | OpenAI: gpt-oss-20bvia PrimeIntellect · fp4 | $0.019/百万缓存读 | input_cache_read: 0.000000019 / token (endpoint primeintellect/fp4) | |
| Aiberm未披露 | MiMo V2 Flash | $0.019/百万缓存读 | model_ratio 0.21 × $2/M · group default ×0.23 · cache_ratio 0.2 | |
| Aiberm未披露 | GPT-5.6 Luna | $0.019/百万缓存读 | model_ratio 0.2826 × $2/M · group default ×0.23 · cache_ratio 0.15 | |
| OpenRouter第三方 | Google: Gemma 3 4Bvia Parasail · bf16 | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint parasail/bf16) | |
| OpenRouter官方 | LiquidAI: LFM3-1.7B | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint liquid/fp8) | |
| OpenRouter官方 | Mistral: Ministral 3 14B 2512via Mistral · zdr | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint mistral/zdr) | |
| OpenRouter官方 | Mistral: Sabavia Mistral · zdr | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint mistral/zdr) | |
| OpenRouter官方合作 | OpenAI: GPT-5.4 Nanovia Azure | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint azure) | |
| OpenRouter官方合作 | OpenAI: GPT-5.6 Lunavia Amazon Bedrock · global | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint amazon-bedrock/global) | |
| OpenRouter官方合作 | OpenAI: GPT-5.6 Luna Provia Azure | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint azure) | |
| OpenRouter官方 | OpenAI: GPT-5.6 Luna Pro (batch)提示词超过 272K | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (>272000; endpoint openai) | |
| OpenRouter官方 | OpenAI: GPT-5.6 Luna (batch)提示词超过 272K | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (>272000; endpoint openai) | |
| OpenRouter官方合作 | OpenAI: GPT-6 Luna提示词超过 272Kvia Amazon Bedrock · global | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (>272000; endpoint amazon-bedrock/global) | |
| OpenRouter官方合作 | OpenAI: GPT-6 Luna Pro提示词超过 272Kvia OpenAI | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (>272000; endpoint openai) | |
| OpenRouter官方 | Qwen: Qwen3.7 Flash提示词超过 32Kvia Alibaba | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (>32000; endpoint alibaba) | |
| OpenRouter第三方 | Z.ai: GLM 5.3 Flash (batch)via Morph · fp8 | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (endpoint morph/fp8) | |
| OpenRouter未披露 | OpenAI: GPT Luna Latest提示词超过 272K | $0.02/百万缓存读 | input_cache_read: 0.00000002 / token (>272000) | |
| Runware未披露 | GPT-5.6 LunaPer 1M | $0.02/百万缓存读 | $2e-8 / cachedInputToken | |
| Runware未披露 | GPT-6 Luna> 272K context | $0.02/百万缓存读 | $2e-8 / cachedInputToken · > 272K context | |
| OpenRouter第三方 | Tencent: Hy3off-peak 1600–0000 UTC · via Tencent · fp8 | $0.021/百万缓存读 | input_cache_read: 0.000000020625 / token (utc_start 1600, utc_end 0; endpoint tencent/fp8) |
