OpenAI 兼容 LLM 网关的意思是：SDK 不动，base_url 指向网关，再从网关路由到各提供方。你可以比单位成本、加缓存或预算，故障切换也不用改应用代码。这页给评估这类接入的团队用：先选靠谱的默认模型，定兜底，再估月费。迁移步骤见 /docs/openai-compatible，成本计算见 /tools/ai-api-cost-calculator。如果主要在比多模型市场，用 /best/openrouter-alternatives。
来源依据: NextModel 目录分类、OpenAI 兼容网关定位，以及可用时的提供方公开定价。 · 更新日期 2026-08-05
步骤
匹配分
先从候选名单开始，再用真实提示词测试，并在接入生产路由前比较月度成本。
This model always redirects to the latest model in the Claude Fable family.
This model always redirects to the latest model in the Anthropic Claude Haiku family.
This model always redirects to the latest model in the Claude Opus family.
This model always redirects to the latest model in the Anthropic Claude Sonnet family.
比较表
缩小生产候选、建立兜底策略或比较模型经济性时用。
| Model | Provider | Input | Output | Context | Capabilities | Best for | Latency | Status | Source |
|---|---|---|---|---|---|---|---|---|---|
| Anthropic: Claude Fable Latest~anthropic/claude-fable-latest | OpenRouter | $1.88 / 1M tokens | $9.40 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| Anthropic Claude Haiku Latest~anthropic/claude-haiku-latest | OpenRouter | $0.188 / 1M tokens | $0.94 / 1M tokens | 200k | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| Anthropic: Claude Opus Latest~anthropic/claude-opus-latest | OpenRouter | $0.94 / 1M tokens | $4.70 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| Anthropic Claude Sonnet Latest~anthropic/claude-sonnet-latest | OpenRouter | $0.564 / 1M tokens | $2.82 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| Google Gemini Flash Latest~google/gemini-flash-latest | OpenRouter | $0.282 / 1M tokens | $1.69 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| Google Gemini Pro Latest~google/gemini-pro-latest | OpenRouter | $0.376 / 1M tokens | $2.26 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| MoonshotAI Kimi Latest~moonshotai/kimi-latest | OpenRouter | $0.124 / 1M tokens | $0.658 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
| OpenAI GPT Latest~openai/gpt-latest | OpenRouter | $0.94 / 1M tokens | $5.64 / 1M tokens | 1.1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1000-3000ms | Catalog | Platform curated |
常见问题
托管 API，接受 OpenAI 请求形态（chat completions、models、keys），在单一 base URL 后转发到一个或多个提供方。
一个集成面、更好做兜底，用量、预算和凭证集中管理，流量变大之前有地方收口。
多数情况改 base_url 和 API key。再用你真正要用的模型 ID 验证 streaming、tools、JSON mode 和 vision。
简单活给便宜模型，难活给更强模型；安全时再缓存；设项目预算。扩容前先估月费。