Model dossier模型档案
GPT-5.6 Sol
GPT-5.6 Sol is a frontier-tier model from OpenAI (Intelligence Score 92.4/100). Strongest in preference (100) and coding (96.3); best fit: general chat and content drafting. Premium-priced: $2.00 in / $10 out per 1M tokens. Context window 1.1M tokens — long-document friendly.
GPT-5.6 Sol 是 OpenAI 的前沿梯队模型(智能评分 92.4/100)。 最强项是偏好(100分),其次是代码(96.3分);适合日常对话与内容起草。 定价属高端定价:每百万 token 输入 $2.00 / 输出 $10。 上下文 1.1M token,长文档友好。
92.4
Intelligence Score · benchmark coverage 智能评分 · benchmark 覆盖 100%
◈
Benchmark scoresBenchmark 成绩
11 entries 条| Benchmark基准 | Score成绩 | Index指数 | Dated日期 | Measured by测评方 | |
|---|---|---|---|---|---|
| Humanity's Last ExamReasoning | 47.2⚡high | 82 | — | 3rd-party第三方 | source ↗×5⚠±6.4 |
| GPQA DiamondReasoning | 94.1⚡max | 99 | — | 3rd-party第三方 | source ↗×7 |
| ↳ ⚡high | 94.1 | — | — | 3rd-party第三方 | source ↗ |
| MMLU-ProKnowledge | 89.1 | 96 | 2026-08 | 3rd-party第三方 | source ↗×2⚠±6.9 |
| SWE-bench VerifiedCoding | 96.2⚡max | 100 | 2026-08 | 3rd-party第三方 | source ↗×7 |
| AIME 2025Math | 97 | 97 | — | 3rd-party第三方 | source ↗×5⚠±3 |
| LiveCodeBenchCoding | 81 | 87 | 2026-07 | 3rd-party第三方 | source ↗×3⚠±3.2 |
| FrontierMathMath | 89.12⚡max | 95 | — | 3rd-party第三方 | source ↗×4⚠±6.12 |
| Terminal-BenchCoding | 91.9 | 100 | — | official官方榜 | source ↗×10⚠±57.3 |
| ↳ ⚡max | 91.9 | — | — | 3rd-party第三方 | source ↗ |
| τ²-benchAgent | 46.9⚡max | 48 | 2026-08-04 | official官方榜 | source ↗×4⚠±38.2 |
| LMArena (Chatbot Arena)Preference | 1623⚡max | 100 | 2026-07-31 | official官方榜 | source ↗×10⚠±142.2 |
| WebArenaAgent | 92.2 | 100 | 2026-07-30 | 3rd-party第三方 | source ↗×2⚠±18.2 |
≡
Specs & pricing规格与价格
| Provider厂商 | OpenAI |
|---|---|
| Released发布 | 2026-07-09 |
| Context window上下文 | 1.1M tokens |
| Max output最大输出 | 128K tokens |
| Modality模态 | text+image+file->text |
| Reasoning tiers推理档位 | ⚡none ⚡low ⚡medium ⚡high ⚡max default默认 high · none / minimal / low / medium / high / xhigh |
| API input price输入价格 | $2.00 / 1M tokens |
| API output price输出价格 | $10 / 1M tokens |
⛓
Tools using this model使用它的工具
⚔
Head-to-head正面对比
How we score →评分方法 → · Catalog & pricing via the official OpenRouter API.模型目录与价格来自 OpenRouter 官方 API。