AI 模型情报

能力 · 2026-08-13

支持结构化输出的 AI 模型

对比支持 JSON mode / 结构化输出的 AI 模型 —— 数据抽取、分类与结构化摘要等管道更稳。

这是什么?

  • 结构化输出(也称 JSON mode 或 response_format=json_schema)将模型约束为你提供的 schema 所匹配的 JSON 文档。
  • 不同于提示里写「请用 JSON 回复」,结构化输出在解码阶段强制约束 —— 模型无法输出非法 JSON。

为什么重要

  • 可避免 JSON 解析错误和「好的,这是 JSON:…」这类越狱前缀。
  • 对任何把 LLM 输出接到类型化系统的流程都至关重要:抽取、分类、结构化摘要等。

506 个模型支持此能力

模型厂商输入 / 1M输出 / 1M上下文服务商
Voxtral Small 24B 2507Mistral$0.002$0.00232K4
Llama 3.2 3B InstructMeta$0.020$0.020131K10
Ling-2.6-flashopenrouter$0.010$0.030262K1
Llama 3.1 8B InstructMeta$0.020$0.030131K18
Mistral Nemo Instruct 2407Mistral$0.020$0.030131K7
Llama 3 8B LunarisMeta$0.040$0.0508K2
Sao10k L3 8B Lunaris novita-ai$0.050$0.0508K1
MythoMax 13Bopenrouter$0.060$0.0608K1
MythoMax 13Bkilo$0.060$0.0604K1
Gemma 4 E4B ITGoogle$0.020$0.100131K3
Nex N2 Mininano-gpt$0.025$0.100262K1
Nex-N2-Miniopenrouter$0.025$0.100262K1
Nex AGI: Nex-N2-Minikilo$0.025$0.100262K1
Qwen3.7 FlashAlibaba (Qwen)$0.030$0.1181M9
Google Gemma 3 12BGoogle$0.050$0.100131K8
Granite 4.1 8Bnano-gpt$0.050$0.100131K1
Granite 4.1 8Bopenrouter$0.050$0.100131K1
IBM: Granite 4.1 8Bkilo$0.050$0.100131K1
Mellum2 12B A2.5Bwandb$0.050$0.100131K1
Granite 4.1 8Bwandb$0.050$0.100131K1
gpt-oss-20bOpenAI$0.030$0.130128K28
GPT OSS 120Bllmgateway$0.032$0.140131K1
Gemma 3n 4BGoogle$0.060$0.12033K4
Qwen3.5 9BAlibaba (Qwen)$0.040$0.150262K22
GPT OSS 20Bllmgateway$0.040$0.150131K1
nova-micro-v1cortecs$0.040$0.159128K1
gpt-oss-120bOpenAI$0.030$0.170128K43
Ministral 3 3B 2512Mistral$0.100$0.100131K3
Reka Edgeopenrouter$0.100$0.10016K1
Ministral 3Bllmgateway$0.100$0.100131K1
Reka Edgekilo$0.100$0.10016K1
GPT OSS 20Bcortecs$0.045$0.167131K1
Mistral-7B-Instruct-v0.3Mistral$0.110$0.11066K5
ministral-3b-2512cortecs$0.111$0.111256K1
Google Gemma 3 27B InstructGoogle$0.080$0.160203K11
Nemotron 3.5 Lightning 30B A3BNVIDIA$0.050$0.2001M3
GPT OSS 20Bdatabricks$0.050$0.200131K1
GPT OSS 20Bneon$0.050$0.200131K1
DeepSeek V4 Flash 0731DeepSeek$0.080$0.1801.05M27
GPT OSS Safeguard 20BOpenAI$0.070$0.200128K6
GPT OSS 20Bfrogbot$0.070$0.200131K1
nvidia-nemotron-3-nano-omniNVIDIA$0.059$0.237300K4
Ministral 3 8B 2512Mistral$0.150$0.150262K3
Ministral 8Bllmgateway$0.150$0.150262K1
Muse Spark 1.2 ContributorMeta$0.100$0.2001.05M3
LFM2.5 2.6Bnano-gpt$0.100$0.200128K1
Reka Flash 3openrouter$0.100$0.20066K1
Reka Flash 3kilo$0.100$0.20066K1
Qwen3.5 FlashAlibaba (Qwen)$0.029$0.2871M7
Tencent Hy3nano-gpt$0.066$0.260262K1
Granite-4.0-H-Smallwatsonx$0.064$0.265131K1
DeepSeek V4 Flash LatestDeepSeek$0.080$0.2521.05M3
ministral-8b-2512cortecs$0.167$0.167256K1
Mistral Small 3.2 24BMistral$0.094$0.250256K3
nova-lite-v1cortecs$0.069$0.275300K1
Nemotron 3.5 Lightning 30B A3BNVIDIA$0.100$0.250262K4
GPT OSS 120Bdatabricks$0.072$0.280131K1
GPT OSS 120Bneon$0.072$0.280131K1
GPT OSS 20Bimpossibl$0.070$0.300131K2
Seed 1.6 Flash (250715)llmgateway$0.070$0.300256K1

显示前 60 / 共 506 项。 完整目录 进一步筛选。

Frequently asked questions

How many AI models support 结构化输出?

506 canonical models in our database currently support 结构化输出. The list is regenerated on every data refresh, so it always reflects the latest releases tracked in our catalogue.

What is the cheapest model with 结构化输出?

Voxtral Small 24B 2507 from Mistral is currently the lowest-priced option, at $0.002 per 1M input tokens and $0.002 per 1M output tokens. The full table above is sorted price-ascending.

Which model with 结构化输出 has the largest context window?

Pokee-Isaac 28B (nano-gpt) leads on context at 10M tokens. This may matter if you also need long-document understanding alongside 结构化输出.

Which models are available on the most providers?

Production-readiness usually correlates with how many independent providers host the same weights. The top three by provider count are: GLM-5.2 (74), Kimi K2.6 (62), DeepSeek V4 Pro (55).

How is 结构化输出 different from a regular LLM?

Structured output (a.k.a. JSON mode / response_format=json_schema) constrains the model at decode time so it cannot emit invalid JSON. This is stricter than just prompting 'reply in JSON' and removes a whole class of parsing errors.

How often is this list updated?

Daily. Our data pipeline syncs once a day, regenerates the canonical model list, and rebuilds these pages so newly released models appear within 24 hours.

最近更新:

Prices in USD per 1M tokens. Unknown means the provider does not publish per-token pricing.

Pricing and capabilities are refreshed daily and reconciled against each provider's official documentation. Always verify critical production decisions with the provider directly.