The lowest sticker price is a weak starting point. Cheap models work well for classification, short summaries, routing, and bulk drafts. They break down when the task needs careful reasoning or long agent runs. Match the model to the job first, then look at input and output rates. The table below carries price, context, capability, and provider source for every candidate — run your numbers through /tools/llm-cost-calculator and /pricing before production.
來源依據: NextModel curated catalog, provider public pricing, and OpenRouter metadata when available. · 更新日期 2026-08-05
如何使用这份名单
综合价格
先從候選名單開始，再以真實提示詞測試，並在接入生產路由前比較月度成本。
FLUX.2 Flex via NextModel: context unpublished; $0.2 / image in / — out per 1M tokens. Capabilities: streaming, vision.
FLUX.2 PRO via NextModel: context unpublished; $0.075 / image in / — out per 1M tokens. Capabilities: streaming, vision.
Doubao Seedance 2.0 via NextModel: context unpublished; $0.148 / s in / — out per 1M tokens. Capabilities: streaming, vision.
Doubao Seedance 2.0 Fast via NextModel: context unpublished; $0.119 / s in / — out per 1M tokens. Capabilities: streaming, vision.
比較表
當你在縮小正式環境候選名單、建立兜底策略或比較模型經濟性時，可使用此視圖。
| Model | Provider | Input | Output | Context | Capabilities | Best for | Latency | Status | Source |
|---|---|---|---|---|---|---|---|---|---|
| FLUX.2 Flexblack-forest-labs/FLUX.2-flex | Black Forest Labs | $0.2 / image | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| FLUX.2 PROblack-forest-labs/FLUX.2-pro | Black Forest Labs | $0.075 / image | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| Doubao Seedance 2.0doubao/doubao-seedance-2-0-260128 | Doubao | $0.148 / s | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| Doubao Seedance 2.0 Fastdoubao/doubao-seedance-2-0-fast-260128 | Doubao | $0.119 / s | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| Doubao Seedance 2.0 minidoubao/doubao-seedance-2-0-mini-260615 | Doubao | $0.0738 / s | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| Doubao Seedance 2.5doubao/doubao-seedance-2-5-260628 | Doubao | $0.225 / s | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| Dreamina-Seedance 2.0dreamina/dreamina-seedance-2-0-260128 | Dreamina | $0.153 / s | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
| Dreamina-Seedance 2.0 Fastdreamina/dreamina-seedance-2-0-fast-260128 | Dreamina | $0.122 / s | — | — | StreamingVision | image understanding, multimodal chat | 0-0ms | Catalog | Platform curated |
常見問題
It depends on FX and how long answers are. In CNY production entries, Doubao Seed 2.0 Mini is usually the lowest here.
No. Use cheap models for repeatable, low-risk work. Keep a stronger model for hard answers, coding agents, and anything you would be embarrassed to ship wrong.
Route easy requests to the cheap model, cache repeated prefixes when it is safe, and keep one OpenAI-compatible base URL. The calculator at /tools/llm-cost-calculator is a quick way to sanity-check the savings.