Long context helps when you stuff contracts, exports, support history, or large files into the prompt. It also makes bills jump. Compare window size and input price together, and decide whether retrieval would be cheaper than stuffing the whole document every time.
來源依據: NextModel curated catalog and OpenRouter context metadata when available. · 更新日期 2026-07-01
如何使用这份名单
上下文长度
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。
GPT-6 Astra via NextModel: 1.1M-token context; Starting at $10 / 1M tokens in / $50 / 1M tokens out per 1M tokens. Capabilities: streaming, tool, json.
GPT-6.1 Sol via NextModel: 1.1M-token context; Starting at $2 / 1M tokens in / $10 / 1M tokens out per 1M tokens. Capabilities: streaming, tool, json.
Claude Fable 5.1 via NextModel: 1M-token context; $10 / 1M tokens in / $50 / 1M tokens out per 1M tokens. Capabilities: streaming, tool, json.
Claude Haiku 5.5 via NextModel: 1M-token context; Starting at $0.1 / 1M tokens in / $0.5 / 1M tokens out per 1M tokens. Capabilities: streaming, tool, json.
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 |
|---|---|---|---|---|---|---|---|---|---|
| GPT-6 Astraopenai/gpt-6-astra | OpenAI | Starting at $10 / 1M tokens | $50 / 1M tokens | 1.1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 2408-7945ms | Catalog | Platform curated |
| GPT-6.1 Solopenai/gpt-6.1-sol | OpenAI | Starting at $2 / 1M tokens | $10 / 1M tokens | 1.1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 2489-8238ms | Catalog | Platform curated |
| Claude Fable 5.1anthropic/claude-fable-5.1 | Anthropic | $10 / 1M tokens | $50 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 3187-5152ms | Catalog | Platform curated |
| Claude Haiku 5.5anthropic/claude-haiku-5.5 | Anthropic | Starting at $0.1 / 1M tokens | $0.5 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1321-2017ms | Catalog | Platform curated |
| Claude Opus 4.6anthropic/claude-opus-4.6 | Anthropic | $5 / 1M tokens | $25 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1930-4195ms | Catalog | Platform curated |
| Claude Opus 5.5anthropic/claude-opus-5.5 | Anthropic | $4 / 1M tokens | $20 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1845-5119ms | Catalog | Platform curated |
| Claude Sonnet 4.6anthropic/claude-sonnet-4.6 | Anthropic | $3 / 1M tokens | $15 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1306-1903ms | Catalog | Platform curated |
| Claude Sonnet 5.5anthropic/claude-sonnet-5.5 | Anthropic | $2 / 1M tokens | $10 / 1M tokens | 1M | StreamingTool callingJSON modeVision | image understanding, multimodal chat | 1230-2392ms | Catalog | Platform curated |
常見問題
No. Bigger windows help with big inputs. Cost, latency, retrieval design, and answer quality still decide whether it is a good idea.
When most of the document is irrelevant to each question. Pull the useful chunks, send less context, and keep a smaller model if quality holds.
Multiply average input tokens (including stuffed documents) by input price, then add output. Long inputs dominate the bill more often than people expect.
相关排行榜