機能 · 2026-05-12
構造化出力に対応した AI モデル
JSON mode / 構造化出力に対応する AI モデルの比較 —— データ抽出や分類パイプライン向け。
これは何か
- 構造化出力（JSON mode や response_format=json_schema など）は、与えたスキーマに一致する JSON のみを出すようモデルを制約します。
- プロンプトで「JSON で答えて」と頼むのとは違い、デコード時に強制されるため、不正な JSON は出せません。
なぜ重要か
- JSON パースエラーや「はい、JSON はこちらです…」のような前置きを減らせます。
- LLM 出力を型付きシステムに流すパイプライン（抽出・分類・構造化要約）に必須です。
この機能に対応するモデル 206 件
全 206 件中、上位 60 件を表示。 さらに絞り込むには モデル一覧 をご利用ください。
Frequently asked questions
How many AI models support 構造化出力?
206 canonical models in our database currently support 構造化出力. The list is regenerated on every data refresh, so it always reflects the latest model releases from models.dev.
What is the cheapest model with 構造化出力?
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 構造化出力 has the largest context window?
GPT-5.4 (OpenAI) leads on context at 1.05M tokens. This may matter if you also need long-document understanding alongside 構造化出力.
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: Kimi K2.5 (45), GPT OSS 120B (33), GLM-5.1 (33).
How is 構造化出力 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 pulls models.dev once a day, regenerates the canonical model list, and rebuilds these pages so newly released models appear within 24 hours.
Explore more
Top models with this capability
- Voxtral Small 24B 2507$0.00 in / $0.00 out
- dots.ocr$0.01 in / $0.01 out
- Hermes 4 14B$0.01 in / $0.05 out
- Sao10k L3 8B Lunaris $0.05 in / $0.05 out
- Gemma 3 12B$0.03 in / $0.10 out
Other capabilities
Best-of lists you might also want
Pricing comparisons
最終更新：
Prices in USD per 1M tokens. Unknown means the provider does not publish per-token pricing.
Data is sourced from models.dev and normalized for comparison. Prices and capabilities may change. Always verify critical production decisions with the provider's official documentation.