Conectá Cursor, Claude Code, OpenCode, VS Code o tus aplicaciones a Anthropic Claude, OpenAI GPT, Google Gemini y DeepSeek. Pagá únicamente por token consumido, sin suscripciones forzadas.
原生兼容支持 OpenAI 或 Anthropic API 的任何工具与 SDK。在下方选择你的环境即可查看对应配置：
使用其他框架？任何兼容 OpenAI Base URL 规范的客户端皆可即插即用。
# 1. Exportá las variables en tu entorno:
export ANTHROPIC_BASE_URL="https://llmpix.com/v1"
export ANTHROPIC_API_KEY="sk-tu_api_key_aqui"
# 2. Ejecutá Claude Code conectado a tu Gateway:
claude --model claude-sonnet-5网关会将请求无缝转换为对应供应商（DeepSeek、Anthropic 或 OpenAI）的原生格式，不修改响应内容，不增加网络延迟。
每一次调用均以原子级精度进行核算：输入 Token、补全 Token 及折合美元扣费一目了然。
Precios reales en USD calculados directamente desde nuestras tarifas de entrada, salida y lectura de caché.
El contexto reutilizado (system prompt, archivos abiertos) se tarifa a precio de lectura de caché.
Cálculo neto según consumo de tokens. Sin cobros sorpresa.
实时同步的美元费率。按模型名称或官方厂商筛选检索。
| 模型名称 | Proveedor Oficial | Input / 1M | Output / 1M | Cache Read / 1M | 服务状态 |
|---|---|---|---|---|---|
| claude-fable-5 | Anthropic | $3.7143 | $18.5714 | $0.3714 | 正常可用 |
| claude-fable-5-1 | Anthropic | $3.7143 | $18.5714 | $0.3714 | 正常可用 |
| claude-fable-5-2 | Anthropic | $2.9714 | $14.8571 | $0.2971 | 正常可用 |
| claude-haiku-4-5-20251001 | Anthropic | $0.1486 | $0.7429 | $0.0149 | 正常可用 |
| claude-haiku-5-5 | Anthropic | $0.1486 | $0.7429 | $0.0149 | 正常可用 |
| claude-opus-4-6 | Anthropic | $1.4857 | $7.4286 | $0.1486 | 正常可用 |
| claude-opus-4-7 | Anthropic | $1.4857 | $7.4286 | $0.1486 | 正常可用 |
| claude-opus-4-8 | Anthropic | $1.4857 | $7.4286 | $0.1486 | 正常可用 |
| claude-opus-5 | Anthropic | $1.4857 | $7.4286 | $0.1486 | 正常可用 |
| claude-opus-5-5 | Anthropic | $1.1886 | $5.9429 | $0.1189 | 正常可用 |
| claude-sonnet-4-6 | Anthropic | $0.2971 | $1.4857 | $0.0297 | 正常可用 |
| claude-sonnet-5 | Anthropic | $0.4457 | $2.2286 | $0.0446 | 正常可用 |
| claude-sonnet-5-5 | Anthropic | $0.5943 | $2.9714 | $0.0594 | 正常可用 |
| deepseek-flash | DeepSeek | $0.2971 | $0.8914 | $0.0446 | 正常可用 |
| deepseek-v4-flash | DeepSeek | $0.2971 | $0.8914 | $0.0446 | 正常可用 |
| deepseek-v4-pro | DeepSeek | $1.3371 | $4.0114 | $0.2006 | 正常可用 |
| doubao-seed-2.1-turbo | ByteDance | $0.4457 | $2.2286 | $0.0557 | 正常可用 |
| gemini-3.7-flash | $0.4457 | $2.6743 | $0.0446 | 正常可用 | |
| gemini-3.8-flash | $0.4457 | $2.6743 | $0.0446 | 正常可用 | |
| gemini-4-argon | $1.4857 | $8.9143 | $0.1486 | 正常可用 | |
| glm-5.1 | Zhipu AI | $0.4457 | $2.2286 | $0.1337 | 正常可用 |
| glm-5.2 | Zhipu AI | $0.7429 | $3.7143 | $0.2229 | 正常可用 |
| glm-5.3 | Zhipu AI | $0.7429 | $3.7143 | $0.2229 | 正常可用 |
| glm-5.3-flash | Zhipu AI | $0.0743 | $0.3714 | $0.0223 | 正常可用 |
| gpt-5.6-luna | OpenAI | $0.0371 | $0.2229 | $0.0046 | 正常可用 |
| gpt-5.6-sol | OpenAI | $0.7429 | $4.4571 | $0.0929 | 正常可用 |
| gpt-5.6-terra | OpenAI | $0.2971 | $1.7829 | $0.0371 | 正常可用 |
| gpt-6-astra | OpenAI | $1.4857 | $8.9143 | $0.1857 | 正常可用 |
| gpt-6-luna | OpenAI | $0.0409 | $0.2455 | $0.0051 | 正常可用 |
| gpt-6-sol | OpenAI | $0.3714 | $2.2286 | $0.0464 | 正常可用 |
| gpt-6.1-sol | OpenAI | $0.3714 | $2.2286 | $0.0464 | 正常可用 |
| gpt-6.1-terra | OpenAI | $0.1857 | $1.1143 | $0.0232 | 正常可用 |
| grok-4.6 | xAI | $0.3714 | $1.1143 | $0.0464 | 正常可用 |
| grok-4.7 | xAI | $0.4457 | $1.3371 | $0.0557 | 正常可用 |
| k3.1-pro | Moonshot AI | $0.8914 | $4.4571 | $0.1114 | 正常可用 |
| k3.1-pro-highspeed | Moonshot AI | $1.7829 | $8.9143 | $0.2229 | 正常可用 |
| kimi-k3 | Moonshot AI | $1.1143 | $5.5714 | $0.1393 | 正常可用 |
| mimo-v2.6-flash | Xiaomi MiMo | $0.1486 | $0.7429 | $0.0149 | 正常可用 |
| mimo-v2.6-pro | Xiaomi MiMo | $0.2971 | $1.4857 | $0.0297 | 正常可用 |
| MiniMax-M3 | MiniMax | $0.0743 | $0.4457 | $0.0186 | 正常可用 |
| MiniMax-M3-highspeed | MiniMax | $0.1486 | $0.8914 | $0.0371 | 正常可用 |
| MiniMax-M3.1-flash | MiniMax | $0.0743 | $0.4457 | $0.0186 | 正常可用 |
| qwen3.7-max | Alibaba Cloud | $0.8914 | $4.4571 | $0.1337 | 正常可用 |
| qwen3.7-plus | Alibaba Cloud | $0.2971 | $1.4857 | $0.0446 | 正常可用 |
| qwen3.8-max | Alibaba Cloud | $1.4857 | $7.4286 | $0.2229 | 正常可用 |
| step-3.7-flash | StepFun | $0.0743 | $0.3714 | $0.0074 | 正常可用 |
Guías técnicas, análisis de costos, integración en IDEs y benchmarks entre proveedores.
Which model should you choose for coding, reasoning, and long context? An objective cost-performance comparison for production workloads.
No existe un único modelo perfecto para todo. Analizamos cuándo conviene utilizar Claude (Anthropic), GPT (OpenAI), Gemini (Google) o DeepSeek para maximizar rendimiento y optimizar presupuesto.
Prompt caching is the quiet revolution in AI engineering. Learn the difference between input, output, and cache hit/write to build architectures that slash your bills to a fraction.
告别繁琐的多币种信用卡、各家账单及不同的 API 协议。仅需单个账户与密钥，畅享市面全部顶尖模型。
支持充值卡与加密货币钱包快速充值。系统根据每条返回严格扣减精确成本。
按需为您不同项目或客户生成专属密钥 (`sk-...`)，随时可一键撤销废除。
在客户后台完整审阅每笔请求明细：输入 Token、输出 Token、调用模型、扣费金额及精准时间戳。
/app 生成的密钥，并将 Base URL 改为 https://llmpix.com/v1。无需额外安装插件，即可立即使用！
402 Payment Required 错误码，充值后即刻恢复。