支持模型与实时费率

Tokenlio 聚合全球前沿大模型与超算集群，统一采用美元标准 ($/1M Tokens) 透明计费。以下按模型研发厂商分类列出。

INFO

如需体验实时延迟测试、模态筛选或快捷键搜索，请访问在线交互式模型矩阵。

OpenAI

Frontier reasoning, coding, and multimodal models.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`gpt-6-astra`	T▧▶◖	1M	$15.00	$45.00	$3.00	20% off
`gpt-5.6-sol`	T▧▶◖	512K	$5.00	$30.00	$2.50	20% off
`gpt-6.1-sol`	T▧▶◖	512K	$12.00	$24.00	$2.40	20% off
`gpt-5.5-pro`	T▧◖	256K	$4.00	$20.00	$1.00	20% off
`gpt-5.4-mini`	T▧	128K	$0.20	$0.80	$0.05	限时免费
`o3-mini`	T	200K	$1.10	$4.40	$0.55	20% off
`gpt-4o`	T▧◖	128K	$2.50	$10.00	$1.25	20% off
`o1`	T▧	200K	$15.00	$60.00	$7.50	20% off
`gpt-4o-mini`	T▧◖	128K	$0.15	$0.60	$0.07	限时免费
`gpt-4o-audio`	T◖	128K	$2.50	$10.00	$1.25	20% off

Anthropic

Long-context reasoning, coding agents, and document analysis.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`claude-opus-5.5`	T▧▶◖	1M	$15.00	$45.00	$1.50	20% off
`claude-fable-5.1`	T▧▶◖	500K	$8.00	$36.00	$0.80	限时 50% off
`claude-sonnet-4.6`	T▧	200K	$3.00	$15.00	$0.30	限时 50% off
`claude-haiku-4.5`	T▧	200K	$0.80	$4.00	$0.08	20% off
`claude-3-opus`	T▧	200K	$15.00	$75.00	$1.50	20% off

DeepSeek

Cost-efficient reasoning, coding, and mathematics models.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`deepseek-v4-pro`	T▧	128K	$0.55	$2.19	$0.11	限时免费
`deepseek-r1`	T	128K	$0.55	$2.19	$0.14	限时免费
`deepseek-chat`	T	128K	$0.14	$0.28	$0.0280	限时免费
`deepseek-coder-v2`	T	128K	$0.14	$0.28	$0.0280	限时 50% off
`deepseek-math-7b`	T	32K	$0.10	$0.20	$0.0200	限时免费

Google

Long-context multimodal models for text, image, video, and audio.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`gemini-3.8-flash`	T▧▶◖	2M	$0.95	$1.90	$0.19	限时 50% off
`gemini-3.1-pro`	T▧▶◖	2M	$1.25	$5.00	$0.31	20% off
`gemini-2.5-pro`	T▧▶◖	2M	$1.25	$5.00	$0.31	限时 50% off
`gemini-2.0-flash`	T▧▶◖	1M	$0.10	$0.40	$0.0250	限时 50% off
`gemini-1.5-pro`	T▧▶◖	2M	$1.25	$5.00	$0.31	20% off
`gemini-2.0-flash-realtime`	T◖▶	1M	$0.15	$0.60	$0.0300	限时 50% off

Alibaba Qwen

Open-weight models for bilingual understanding, coding, and vision.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`qwen3.8-max`	T▧	128K	$1.60	$4.80	$0.32	20% off
`qwen3.7-plus`	T▧	128K	$0.80	$2.40	$0.16	限时 50% off
`qwen-2.5-coder-32b`	T	128K	$0.18	$0.36	$0.0360	限时免费
`qvq-72b-preview`	T▧	32K	$0.80	$2.40	$0.16	限时 50% off
`qwen-2.5-72b-instruct`	T	128K	$0.35	$0.40	$0.07	限时免费
`qwen-2-audio-7b`	T◖	32K	$0.15	$0.30	$0.0300	限时免费

Moonshot Kimi

Long-context understanding and deep search models.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`kimi-k3`	T▧	200K+	$1.80	$3.60	$0.36	限时 50% off
`kimi-k3-free`	T	128K	$0.000	$0.000	$0.0000	限时免费

xAI

High-performance real-world knowledge and reasoning models.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`grok-4.5`	T▧	256K	$2.00	$6.00	$0.40	限时 50% off
`grok-4.3`	T▧	128K	$2.00	$8.00	$0.40	20% off
`grok-vision-beta`	T▧	128K	$5.00	$15.00	$1.00	20% off

Zhipu GLM

Bilingual models with vision and multi-agent capabilities.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`glm-5.3`	T▧▶	128K	$1.20	$2.40	$0.24	20% off
`glm-5.2`	T▧	128K	$1.00	$2.00	$0.20	20% off
`glm-4-flash`	T	128K	$0.060	$0.060	$0.0120	限时免费
`glm-4v-plus`	T▧▶	128K	$1.40	$2.80	$0.28	限时 50% off
`glm-4-long`	T	1M	$0.50	$1.00	$0.10	限时 50% off

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`llama-3.3-70b-instruct`	T	128K	$0.35	$0.40	$0.07	限时免费
`llama-3.1-405b-instruct`	T	128K	$1.80	$3.60	$0.36	20% off
`llama-3.2-90b-vision`	T▧	128K	$0.90	$0.90	$0.18	限时 50% off
`llama-3.2-11b-vision`	T▧	128K	$0.15	$0.15	$0.0300	限时 50% off

Mistral

European frontier, coding, and vision models.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`mistral-large-2411`	T	128K	$2.00	$6.00	$0.40	限时 50% off
`codestral-2501`	T	256K	$0.30	$0.90	$0.06	20% off
`pixtral-12b`	T▧	128K	$0.15	$0.15	$0.0300	限时免费
`pixtral-large-124b`	T▧	128K	$2.00	$6.00	$0.40	20% off

MiniMax

Ultra-long-context and multimodal models at efficient pricing.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`minimax-m3`	T◖	1M	$0.20	$1.10	$0.0400	限时 50% off
`minimax-01`	T▧	4M	$0.20	$1.10	$0.0400	限时 50% off
`minimax-text-01`	T	1M	$0.15	$0.90	$0.0300	限时免费
`minimax-video-01`	T▶	32K	$1.50	$4.50	$0.30	20% off

Cohere

Enterprise RAG, search, and grounded generation models.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`command-r-plus`	T	128K	$2.50	$10.00	$0.50	20% off
`command-r`	T	128K	$0.50	$1.50	$0.10	限时 50% off

NVIDIA

Hardware-accelerated models for high-throughput agent workloads.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`nemotron-3-nano-omni`	T▧◖	128K	$0.080	$0.16	$0.0160	限时免费

Amazon

AWS-native Nova models for reliable enterprise workloads.

模型名称	支持模态	上下文	输入价格 / 1M	输出价格 / 1M	缓存读 / 1M	优惠活动
`nova-pro-v1`	T▧▶	300K	$0.80	$3.20	$0.20	20% off
`nova-lite-v1`	T▧▶	300K	$0.060	$0.24	$0.0150	限时 50% off
`nova-micro-v1`	T	128K	$0.035	$0.14	$0.0088	限时免费

快速调用示例

bash

curl https://api.tokenlio.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "openai/gpt-6-astra",
    "messages": [{"role": "user", "content": "Hello Tokenlio!"}]
  }'

支持模型与实时费率 ​

OpenAI ​

Anthropic ​

DeepSeek ​

Google ​

Alibaba Qwen ​

Moonshot Kimi ​

xAI ​

Zhipu GLM ​

Meta ​

Mistral ​

MiniMax ​

Cohere ​

NVIDIA ​

Amazon ​

快速调用示例 ​