AIモデルインテリジェンス

機能 · 2026-05-12

画像入力に対応した AI モデル

画像とテキストを一緒に入力できる AI モデルの比較。

これは何か

  • ビジョン言語モデルはテキストに加え(または代わりに)画像入力を受け付けます。
  • 多くはテキスト入出力のマルチモーダル LLM であり、画像生成モデルではありません。

なぜ重要か

  • 用途例: 文書理解(スキャン、PDF、スクリーンショット)、UI/コードのスクリーンショットレビュー、商品写真 Q&A、アクセシビリティ(alt)、医療/衛星画像など。
  • 料金は画像枚数に加えトークン課金が乗ることが多いです —— 各プロバイダーの offering 表を確認してください。

この機能に対応するモデル 382 件

モデルベンダー入力 / 1M出力 / 1Mコンテキストプロバイダー
dots.ocrchutes$0.010$0.011131K1
Gemma 3 4BGoogle$0.010$0.02733K7
PaddleOCR-VLnovita-ai$0.020$0.02016K1
Llama-3.2-11B-Vision-InstructMeta$0.049$0.049128K8
Gemma 3 12BGoogle$0.030$0.10033K10
Gemma 3 27BGoogle$0.027$0.109131K14
Model Routerazure-cognitive-services$0.140Unknown128K1
Model Routerazure$0.140Unknown128K1
Gemini 1.5 Flash-8BGoogle$0.037$0.1501M1
Qwen/Qwen3.5-9BAlibaba (Qwen)$0.050$0.150262K6
Qwen/Qwen3-VL-30B-A3B-ThinkingAlibaba (Qwen)$0.100$0.100262K6
Qwen/Qwen3-VL-30B-A3B-InstructAlibaba (Qwen)$0.100$0.100262K6
Qwen/Qwen3-VL-8B-InstructAlibaba (Qwen)$0.100$0.100262K5
Reka Edgekilo$0.100$0.10016K1
Ministral 3Bllmgateway$0.100$0.100131K1
GLM-4.6V-FlashZ.AI / Zhipu$0.020$0.210128K3
Mistral Small 3.2 24B InstructMistral$0.060$0.18096K3
Qwen/Qwen2.5-VL-32B-InstructAlibaba (Qwen)$0.050$0.220131K6
Pixtral 12BMistral$0.150$0.150128K2
Amazon: Nova Lite 1.0kilo$0.060$0.240300K1
Nova Liteamazon-bedrock$0.060$0.240300K1
Nova Litevercel$0.060$0.240300K1
Ministral 8Bllmgateway$0.150$0.150262K1
Llama 3.2 11B Vision InstructMeta$0.160$0.160128K1
Qwen-Omni TurboAlibaba (Qwen)$0.070$0.27033K3
Llama Guard 4 12BMeta$0.180$0.180131K3
Arcee AI: Spotlightkilo$0.180$0.180131K1
Seed 1.6 Flash (250715)llmgateway$0.070$0.300256K1
Gemini 2.0 Flash LiteGoogle$0.075$0.3001.05M8
ByteDance Seed: Seed 1.6 Flashkilo$0.075$0.300262K1
Gemini 1.5 FlashGoogle$0.075$0.3001M1
Llama 4 Scout 17B 16E InstructMeta$0.080$0.300128K12
Llama-4-Scout-17B-16E-Instruct-FP8Meta$0.080$0.300128K5
Gemma 4 26BGoogle$0.100$0.300256K8
Mistral Small 3.1Mistral$0.100$0.300128K3
Phi-4-multimodalMicrosoft$0.080$0.320128K2
Mistral Small 3.2Mistral$0.100$0.300128K2
Pixtral 12B 2409scaleway$0.200$0.200128K1
Molmo 2 8Bnano-gpt$0.200$0.20037K1
Ministral 14Bllmgateway$0.200$0.200262K1
Mistral Small 3.2 24B Instruct (2506)Mistral$0.100$0.310128K5
Meta Llama Guard 4 12BMeta$0.210$0.210131K1
GPT-5 NanoOpenAI$0.050$0.400400K17
Kilo Auto Smallkilo$0.050$0.400400K1
Gemini 2.5 Flash LiteGoogle$0.100$0.4001.05M13
GPT-4.1 nanoOpenAI$0.100$0.4001.05M12
Gemini 2.5 Flash Lite Preview 09-25Google$0.100$0.4001.05M9
Gemini 2.0 FlashGoogle$0.100$0.4001.05M6
Gemini 2.5 Flash Lite Preview 06-17Google$0.100$0.4001.05M4
Gemini 2.0 FlashGoogle$0.100$0.4001.05M3
Qwen/Qwen3-Omni-30B-A3B-ThinkingAlibaba (Qwen)$0.100$0.40066K3
Qwen/Qwen3-Omni-30B-A3B-InstructAlibaba (Qwen)$0.100$0.40066K3
Qwen3.5 FlashAlibaba (Qwen)$0.100$0.4001M3
Qwen2.5-Omni 7BAlibaba (Qwen)$0.100$0.40033K2
Gemini Flash-Lite LatestGoogle$0.100$0.4001.05M2
ByteDance Seed: Seed-2.0-Minikilo$0.100$0.400262K1
Gemma 4 31BGoogle$0.130$0.380256K11
Qwen/Qwen3-VL-32B-InstructAlibaba (Qwen)$0.104$0.416262K3
Grok 4.1 Fast (Reasoning)xAI$0.180$0.450128K10
Grok 4 Fast (Reasoning)xAI$0.180$0.4502M9

全 382 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 画像入力?

382 canonical models in our database currently support 画像入力. 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 画像入力?

dots.ocr from chutes is currently the lowest-priced option, at $0.010 per 1M input tokens and $0.011 per 1M output tokens. The full table above is sorted price-ascending.

Which model with 画像入力 has the largest context window?

Llama 4 Scout 17B Instruct (Meta) leads on context at 3.50M tokens. This may matter if you also need long-document understanding alongside 画像入力.

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), Kimi K2.6 (31), Qwen3.5 397B-A17B (22).

How is 画像入力 different from a regular LLM?

Vision-language models accept image input alongside text. They are multimodal LLMs, not image generators — most reply in text after looking at the image.

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.

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.