---
title: Utilizzo della Messages API
url: https://platform.claude.com/docs/it/build-with-claude/working-with-messages
description: Pattern pratici ed esempi per utilizzare l'API Messages in modo efficace
featureMetadata:
  status: stable
  zdr:
    eligibility: eligible
    note: Excludes [Covered Models](https://platform.claude.com/docs/en/manage-claude/api-and-data-retention#model-specific-data-retention-requirements). Chiamate API standard per generare le risposte di Claude.
  supportedModels:
    - claude-fable-5-2
    - claude-mythos-5-2
    - claude-fable-5-1
    - claude-mythos-5-1
    - claude-fable-5
    - claude-mythos-5
    - claude-mythos-preview
    - claude-opus-5-5
    - claude-opus-5
    - claude-sonnet-5-5
    - claude-sonnet-5
    - claude-haiku-5-5
    - claude-haiku-4-5-20251001
    - claude-opus-4-8
    - claude-opus-4-7
    - claude-opus-4-6
    - claude-sonnet-4-6
    - claude-opus-4-5-20251101
    - claude-sonnet-4-5-20250929
  supportedPlatforms:
    Claude API: stable
    Amazon Bedrock: stable
    Amazon Bedrock (bedrock-runtime): stable
    Claude Platform on AWS: stable
    Claude Models on Google Cloud (Anthropic managed): stable
    Claude Models on Google Cloud (Google managed): stable
    Microsoft Foundry: stable
---

Anthropic offre due modi per sviluppare con Claude, ciascuno adatto a casi d’uso diversi:

|                | Messages API                                        | Claude Managed Agents                                                                                |
| -------------- | --------------------------------------------------- | ---------------------------------------------------------------------------------------------------- |
| **Cos’è**      | Accesso diretto al prompting del modello            | Harness per agenti preconfigurato e personalizzabile che viene eseguito su un’infrastruttura gestita |
| **Ideale per** | Loop di agenti personalizzati e controllo granulare | Attività di lunga durata e lavoro asincrono                                                          |

Questa guida tratta i pattern comuni per lavorare con l’API Messages, incluse le richieste di base, le conversazioni a più turni, le tecniche di “prefill” (precompilazione) e le capacità di visione. Per le specifiche complete dell’API, consulta il [riferimento dell’API Messages](https://platform.claude.com/docs/it/api/messages/create). Per l’harness di agenti gestito, consulta invece la [panoramica di Claude Managed Agents](https://platform.claude.com/docs/it/managed-agents/overview).

## Richiesta e risposta di base

<Note>
  I parametri di campionamento `temperature`, `top_p` e `top_k` non sono supportati su Claude 4.7 e sui modelli successivi e su Claude Mythos Preview, né su alcun modello a partire dalla [versione API `2026-10-21`](https://platform.claude.com/docs/it/api/migration-2026-10-21#sampling-parameters). Impostarli su un valore diverso da quello predefinito restituisce un errore 400, e a partire da `2026-10-21` lo fa qualsiasi valore. Omettili dai payload delle richieste e usa invece il prompting per guidare il comportamento del modello. Consulta le [modifiche che causano incompatibilità nella guida alla migrazione](https://platform.claude.com/docs/it/models/opus-5-5/migration-guide#opus-46-breaking-changes).
</Note>

<CodeGroup>
  ```bash cURL
  #!/bin/sh
  curl https://api.anthropic.com/v1/messages \
    -H "x-api-key: $ANTHROPIC_API_KEY" \
    -H "anthropic-version: 2026-10-21" \
    -H "content-type: application/json" \
    -d '{
      "model": "claude-opus-5-5",
      "max_tokens": 1024,
      "messages": [
        {"role": "user", "content": "Hello, Claude"}
      ]
    }'
  ```

  ```bash CLI
  ant messages create \
    --model claude-opus-5-5 \
    --max-tokens 1024 \
    --message '{role: user, content: "Hello, Claude"}'
  ```

  ```python Python
  message = anthropic.Anthropic().messages.create(
      model="claude-opus-5-5",
      max_tokens=1024,
      messages=[{"role": "user", "content": "Hello, Claude"}],
  )
  print(message)
  ```

  ```typescript TypeScript
  const anthropic = new Anthropic();

  const message = await anthropic.messages.create({
    model: "claude-opus-5-5",
    max_tokens: 1024,
    messages: [{ role: "user", content: "Hello, Claude" }]
  });
  console.log(message);
  ```

  ```csharp C#
  AnthropicClient client = new();

  var parameters = new MessageCreateParams
  {
      Model = Model.ClaudeOpus5_5,
      MaxTokens = 1024,
      Messages = [new UserMessageParam { Content = "Hello, Claude" }]
  };
  var message = await client.Messages.Create(parameters);
  Console.WriteLine(message);
  ```

  ```go Go
  client := anthropic.NewClient()

  response, err := client.Messages.New(context.TODO(), anthropic.MessageNewParams{
  	Model:     anthropic.ModelClaudeOpus5_5,
  	MaxTokens: 1024,
  	Messages: []anthropic.MessageParam{
  		anthropic.NewUserMessage(anthropic.NewTextBlock("Hello, Claude")),
  	},
  })
  if err != nil {
  	log.Fatal(err)
  }
  fmt.Println(response)
  ```

  ```java Java
  AnthropicClient client = AnthropicOkHttpClient.fromEnv();

  MessageCreateParams params = MessageCreateParams.builder()
      .model(Model.CLAUDE_OPUS_5_5)
      .maxTokens(1024L)
      .addUserMessage("Hello, Claude")
      .build();

  Message response = client.messages().create(params);
  System.out.println(response);
  ```

  ```php PHP
  $client = new Client();

  $message = $client->messages->create(
      maxTokens: 1024,
      messages: [['role' => 'user', 'content' => 'Hello, Claude']],
      model: Model::CLAUDE_OPUS_5_5,
  );
  echo json_encode($message, JSON_PRETTY_PRINT), PHP_EOL;
  ```

  ```ruby Ruby
  client = Anthropic::Client.new

  message = client.messages.create(
    model: Anthropic::Model::CLAUDE_OPUS_5_5,
    max_tokens: 1024,
    messages: [
      { role: "user", content: "Hello, Claude" }
    ]
  )
  puts message
  ```
</CodeGroup>

```json Output
{
  "id": "msg_01XFDUDYJgAACzvnptvVoYEL",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "Hello!"
    }
  ],
  "model": "claude-opus-5-5",
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 12,
    "output_tokens": 6
  }
}
```

Le risposte di rifiuto (`stop_reason: "refusal"`) includono anche un oggetto `stop_details` che identifica la categoria di policy che ha attivato il rifiuto, su ogni modello. Consulta [Come appare un rifiuto](https://platform.claude.com/docs/it/build-with-claude/refusals-and-fallback#refusal-response) per il riferimento dei campi e codice di esempio per la gestione.

## Turni conversazionali multipli

L’API Messages è stateless, il che significa che invii sempre l’intera cronologia conversazionale all’API. Puoi usare questo pattern per costruire una conversazione nel tempo. I turni conversazionali precedenti non devono necessariamente provenire effettivamente da Claude. Puoi usare messaggi `assistant` sintetici.

Con i [thread di messaggi](https://platform.claude.com/docs/it/build-with-claude/message-threads), in beta, l’API può conservare la conversazione, in modo che ogni richiesta successiva invii solo i nuovi messaggi. Un thread scade 5 minuti dopo la sua ultima richiesta e viene eliminato non oltre 1 ora dopo il suo ultimo utilizzo.

<CodeGroup>
  ```bash cURL
  #!/bin/sh
  curl https://api.anthropic.com/v1/messages \
    -H "x-api-key: $ANTHROPIC_API_KEY" \
    -H "anthropic-version: 2026-10-21" \
    -H "content-type: application/json" \
    -d '{
      "model": "claude-opus-5-5",
      "max_tokens": 1024,
      "messages": [
        {"role": "user", "content": "Hello, Claude"},
        {"role": "assistant", "content": "Hello!"},
        {"role": "user", "content": "Can you describe LLMs to me?"}

      ]
    }'
  ```

  ```bash CLI
  ant messages create \
    --model claude-opus-5-5 \
    --max-tokens 1024 \
    --message '{role: user, content: "Hello, Claude"}' \
    --message '{role: assistant, content: "Hello!"}' \
    --message '{role: user, content: "Can you describe LLMs to me?"}'
  ```

  ```python Python
  message = anthropic.Anthropic().messages.create(
      model="claude-opus-5-5",
      max_tokens=1024,
      messages=[
          {"role": "user", "content": "Hello, Claude"},
          {"role": "assistant", "content": "Hello!"},
          {"role": "user", "content": "Can you describe LLMs to me?"},
      ],
  )
  print(message)
  ```

  ```typescript TypeScript
  const anthropic = new Anthropic();

  const message = await anthropic.messages.create({
    model: "claude-opus-5-5",
    max_tokens: 1024,
    messages: [
      { role: "user", content: "Hello, Claude" },
      { role: "assistant", content: "Hello!" },
      { role: "user", content: "Can you describe LLMs to me?" }
    ]
  });
  console.log(message);
  ```

  ```csharp C#
  AnthropicClient client = new();

  var parameters = new MessageCreateParams
  {
      Model = Model.ClaudeOpus5_5,
      MaxTokens = 1024,
      Messages =
      [
          new UserMessageParam { Content = "Hello, Claude" },
          new AssistantMessageParam { Content = "Hello!" },
          new UserMessageParam { Content = "Can you describe LLMs to me?" }
      ]
  };

  var message = await client.Messages.Create(parameters);
  Console.WriteLine(message);
  ```

  ```go Go
  client := anthropic.NewClient()

  response, err := client.Messages.New(context.TODO(), anthropic.MessageNewParams{
  	Model:     anthropic.ModelClaudeOpus5_5,
  	MaxTokens: 1024,
  	Messages: []anthropic.MessageParam{
  		anthropic.NewUserMessage(anthropic.NewTextBlock("Hello, Claude")),
  		anthropic.NewAssistantMessage(anthropic.NewTextBlock("Hello!")),
  		anthropic.NewUserMessage(anthropic.NewTextBlock("Can you describe LLMs to me?")),
  	},
  })
  if err != nil {
  	log.Fatal(err)
  }
  fmt.Println(response)
  ```

  ```java Java
  AnthropicClient client = AnthropicOkHttpClient.fromEnv();

  MessageCreateParams params = MessageCreateParams.builder()
      .model(Model.CLAUDE_OPUS_5_5)
      .maxTokens(1024L)
      .addUserMessage("Hello, Claude")
      .addAssistantMessage("Hello!")
      .addUserMessage("Can you describe LLMs to me?")
      .build();

  Message response = client.messages().create(params);
  System.out.println(response);
  ```

  ```php PHP
  $client = new Client();

  $message = $client->messages->create(
      maxTokens: 1024,
      messages: [
          ['role' => 'user', 'content' => 'Hello, Claude'],
          ['role' => 'assistant', 'content' => 'Hello!'],
          ['role' => 'user', 'content' => 'Can you describe LLMs to me?'],
      ],
      model: Model::CLAUDE_OPUS_5_5,
  );

  echo json_encode($message, JSON_PRETTY_PRINT), PHP_EOL;
  ```

  ```ruby Ruby
  client = Anthropic::Client.new

  message = client.messages.create(
    model: Anthropic::Model::CLAUDE_OPUS_5_5,
    max_tokens: 1024,
    messages: [
      { role: "user", content: "Hello, Claude" },
      { role: "assistant", content: "Hello!" },
      { role: "user", content: "Can you describe LLMs to me?" }
    ]
  )
  puts message
  ```
</CodeGroup>

```json Output
{
  "id": "msg_018gCsTGsXkYJVqYPxTgDHBU",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "Sure, I'd be happy to provide..."
    }
  ],
  "model": "claude-opus-5-5",
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 30,
    "output_tokens": 309
  }
}
```

### Ruolo di sistema nei messaggi

Per aggiungere una nuova istruzione di sistema a metà di una conversazione, includi un messaggio con `"role": "system"` subito dopo un turno dell’utente. Quel messaggio non può essere la prima voce in `messages`. Questo funziona su Claude Fable 5.2, [Claude Mythos 5.2](https://platform.claude.com/docs/it/models/mythos-5-2/overview), Claude Fable 5.1, [Claude Mythos 5.1](https://platform.claude.com/docs/it/models/mythos-5-1/overview), Claude Fable 5, [Claude Mythos 5](https://platform.claude.com/docs/it/models/mythos-5/overview), Claude Opus 5.5, Claude Opus 5, Claude Opus 4.8, Claude Sonnet 5.5 e Claude Haiku 5.5. Consulta le [regole di posizionamento](https://platform.claude.com/docs/it/build-with-claude/mid-conversation-system-messages#limitations) complete.

Un messaggio di sistema a metà conversazione ha la stessa autorità del campo `system` di primo livello, ma poiché viene aggiunto alla fine della cronologia dei messaggi, non invalida alcun prefisso memorizzato nella cache che lo precede. Usa il campo `system` di primo livello per le istruzioni che devono applicarsi fin dal primo turno, e un messaggio di sistema a metà conversazione per le istruzioni che diventano rilevanti solo in seguito.

Consulta [Messaggi di sistema e modifiche agli strumenti a metà conversazione](https://platform.claude.com/docs/it/build-with-claude/mid-conversation-system-messages) per la guida completa, incluso come combinarlo con la [“prompt caching” (cache dei prompt)](https://platform.claude.com/docs/it/build-with-claude/prompt-caching).

## Precompilazione della risposta di Claude

Puoi precompilare parte della risposta di Claude nell’ultima posizione dell’elenco dei messaggi di input. Usa questa tecnica per modellare la risposta di Claude. L’esempio seguente usa `"max_tokens": 1` per ottenere da Claude una singola risposta a scelta multipla.

<Warning>
  La precompilazione non è supportata su Claude 4.6 e modelli successivi e su [Claude Mythos Preview](https://anthropic.com/glasswing). Le richieste che usano la precompilazione con questi modelli restituiscono un errore 400. Usa invece gli [output strutturati](https://platform.claude.com/docs/it/build-with-claude/structured-outputs) sui modelli che li supportano, oppure istruzioni nel prompt di sistema. Consulta [Guide alla migrazione](https://platform.claude.com/docs/it/about-claude/models/migration-guide) per i pattern di migrazione.
</Warning>

<CodeGroup>
  ```bash cURL
  #!/bin/sh
  curl https://api.anthropic.com/v1/messages \
    -H "x-api-key: $ANTHROPIC_API_KEY" \
    -H "anthropic-version: 2026-10-21" \
    -H "content-type: application/json" \
    -d '{
      "model": "claude-sonnet-4-5",
      "max_tokens": 1,
      "messages": [
        {"role": "user", "content": "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"},
        {"role": "assistant", "content": "The answer is ("}
      ]
    }'
  ```

  ```bash CLI
  ant messages create <<'YAML'
  model: claude-sonnet-4-5
  max_tokens: 1
  messages:
    - role: user
      content: "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"
    - role: assistant
      content: "The answer is ("
  YAML
  ```

  ```python Python
  message = anthropic.Anthropic().messages.create(
      model="claude-sonnet-4-5",
      max_tokens=1,
      messages=[
          {
              "role": "user",
              "content": "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae",
          },
          {"role": "assistant", "content": "The answer is ("},
      ],
  )
  print(message)
  ```

  ```typescript TypeScript
  const anthropic = new Anthropic();

  const message = await anthropic.messages.create({
    model: "claude-sonnet-4-5",
    max_tokens: 1,
    messages: [
      {
        role: "user",
        content: "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"
      },
      { role: "assistant", content: "The answer is (" }
    ]
  });
  console.log(message);
  ```

  ```csharp C#
  AnthropicClient client = new();

  var parameters = new MessageCreateParams
  {
      Model = Model.ClaudeSonnet4_5,
      MaxTokens = 1,
      Messages = [
          new UserMessageParam { Content = "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae" },
          new AssistantMessageParam { Content = "The answer is (" }
      ]
  };

  var message = await client.Messages.Create(parameters);
  Console.WriteLine(message);
  ```

  ```go Go
  client := anthropic.NewClient()

  response, err := client.Messages.New(context.TODO(), anthropic.MessageNewParams{
  	Model:     anthropic.ModelClaudeSonnet4_5,
  	MaxTokens: 1,
  	Messages: []anthropic.MessageParam{
  		anthropic.NewUserMessage(anthropic.NewTextBlock("What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae")),
  		anthropic.NewAssistantMessage(anthropic.NewTextBlock("The answer is (")),
  	},
  })
  if err != nil {
  	log.Fatal(err)
  }
  fmt.Println(response)
  ```

  ```java Java
  AnthropicClient client = AnthropicOkHttpClient.fromEnv();

  MessageCreateParams params = MessageCreateParams.builder()
      .model(Model.CLAUDE_SONNET_4_5)
      .maxTokens(1L)
      .addUserMessage("What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae")
      .addAssistantMessage("The answer is (")
      .build();

  Message response = client.messages().create(params);
  System.out.println(response);
  ```

  ```php PHP
  $client = new Client();

  $message = $client->messages->create(
      maxTokens: 1,
      messages: [
          ['role' => 'user', 'content' => 'What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae'],
          ['role' => 'assistant', 'content' => 'The answer is ('],
      ],
      model: 'claude-sonnet-4-5',
  );
  echo $message->content[0]->text;
  ```

  ```ruby Ruby
  client = Anthropic::Client.new

  message = client.messages.create(
    model: Anthropic::Model::CLAUDE_SONNET_4_5,
    max_tokens: 1,
    messages: [
      {
        role: "user",
        content: "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"
      },
      { role: "assistant", content: "The answer is (" }
    ]
  )
  puts message
  ```
</CodeGroup>

```json Output
{
  "id": "msg_01Q8Faay6S7QPTvEUUQARt7h",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "C"
    }
  ],
  "model": "claude-sonnet-4-5",
  "stop_reason": "max_tokens",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 42,
    "output_tokens": 1
  }
}
```

## Visione

Claude può leggere sia testo che immagini nelle richieste. Puoi fornire immagini usando i tipi di sorgente `base64`, `url` o `file`. Il tipo di sorgente `file` fa riferimento a un’immagine caricata tramite la [Files API](https://platform.claude.com/docs/it/build-with-claude/files). I tipi di media supportati sono `image/jpeg`, `image/png`, `image/gif` e `image/webp`. Consulta la guida [Visione](https://platform.claude.com/docs/it/build-with-claude/vision) per maggiori dettagli.

<CodeGroup>
  ```bash cURL
  #!/bin/sh

  # Opzione 1: immagine codificata in base64
  IMAGE_BASE64=$(curl "https://platform.claude.com/docs/images/vision-example.jpg" | base64 | tr -d '\n')

  curl https://api.anthropic.com/v1/messages \
    -H "x-api-key: $ANTHROPIC_API_KEY" \
    -H "anthropic-version: 2026-10-21" \
    -H "content-type: application/json" \
    -d @- <<EOF
  {
    "model": "claude-opus-5-5",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": [
        {"type": "image", "source": {
          "type": "base64",
          "media_type": "image/jpeg",
          "data": "$IMAGE_BASE64"
        }},
        {"type": "text", "text": "What is in the above image?"}
      ]}
    ]
  }
  EOF

  # Opzione 2: immagine referenziata tramite URL
  curl https://api.anthropic.com/v1/messages \
    -H "x-api-key: $ANTHROPIC_API_KEY" \
    -H "anthropic-version: 2026-10-21" \
    -H "content-type: application/json" \
    -d '{
      "model": "claude-opus-5-5",
      "max_tokens": 1024,
      "messages": [
        {"role": "user", "content": [
          {"type": "image", "source": {
            "type": "url",
            "url": "https://platform.claude.com/docs/images/vision-example.jpg"
          }},
          {"type": "text", "text": "What is in the above image?"}
        ]}
      ]
    }'
  ```

  ```bash CLI
  # Opzione 1: immagine codificata in base64 (la CLI codifica in automatico i riferimenti binari @file)
  curl -s "https://platform.claude.com/docs/images/vision-example.jpg" \
    -o ./vision-example.jpg

  ant messages create <<'YAML'
  model: claude-opus-5-5
  max_tokens: 1024
  messages:
    - role: user
      content:
        - type: image
          source:
            type: base64
            media_type: image/jpeg
            data: "@./vision-example.jpg"
        - type: text
          text: What is in the above image?
  YAML

  # Opzione 2: immagine referenziata tramite URL
  ant messages create <<'YAML'
  model: claude-opus-5-5
  max_tokens: 1024
  messages:
    - role: user
      content:
        - type: image
          source:
            type: url
            url: https://platform.claude.com/docs/images/vision-example.jpg
        - type: text
          text: What is in the above image?
  YAML
  ```

  ```python Python
  import base64
  import httpx2

  # Opzione 1: immagine codificata in base64
  image_url = "https://platform.claude.com/docs/images/vision-example.jpg"
  image_media_type = "image/jpeg"
  image_data = base64.standard_b64encode(httpx2.get(image_url).content).decode("utf-8")

  message = anthropic.Anthropic().messages.create(
      model="claude-opus-5-5",
      max_tokens=1024,
      messages=[
          {
              "role": "user",
              "content": [
                  {
                      "type": "image",
                      "source": {
                          "type": "base64",
                          "media_type": image_media_type,
                          "data": image_data,
                      },
                  },
                  {"type": "text", "text": "What is in the above image?"},
              ],
          }
      ],
  )
  print(message)

  # Opzione 2: immagine referenziata tramite URL
  message_from_url = anthropic.Anthropic().messages.create(
      model="claude-opus-5-5",
      max_tokens=1024,
      messages=[
          {
              "role": "user",
              "content": [
                  {
                      "type": "image",
                      "source": {
                          "type": "url",
                          "url": "https://platform.claude.com/docs/images/vision-example.jpg",
                      },
                  },
                  {"type": "text", "text": "What is in the above image?"},
              ],
          }
      ],
  )
  print(message_from_url)
  ```

  ```typescript TypeScript
  const anthropic = new Anthropic();

  // Opzione 1: immagine codificata in base64
  const imageUrl = "https://platform.claude.com/docs/images/vision-example.jpg";
  const imageMediaType = "image/jpeg";
  const imageArrayBuffer = await (await fetch(imageUrl)).arrayBuffer();
  const imageData = Buffer.from(imageArrayBuffer).toString("base64");

  const message = await anthropic.messages.create({
    model: "claude-opus-5-5",
    max_tokens: 1024,
    messages: [
      {
        role: "user",
        content: [
          {
            type: "image",
            source: {
              type: "base64",
              media_type: imageMediaType,
              data: imageData
            }
          },
          {
            type: "text",
            text: "What is in the above image?"
          }
        ]
      }
    ]
  });
  console.log(message);

  // Opzione 2: immagine referenziata tramite URL
  const messageFromUrl = await anthropic.messages.create({
    model: "claude-opus-5-5",
    max_tokens: 1024,
    messages: [
      {
        role: "user",
        content: [
          {
            type: "image",
            source: {
              type: "url",
              url: "https://platform.claude.com/docs/images/vision-example.jpg"
            }
          },
          {
            type: "text",
            text: "What is in the above image?"
          }
        ]
      }
    ]
  });
  console.log(messageFromUrl);
  ```

  ```csharp C#
  using System.Collections.Generic;
  using System.Net.Http;
  using Anthropic;
  using Anthropic.Models.Messages;

  AnthropicClient client = new();

  // Opzione 1: immagine codificata in base64
  string imageUrl = "https://platform.claude.com/docs/images/vision-example.jpg";

  using HttpClient httpClient = new();
  byte[] imageBytes = await httpClient.GetByteArrayAsync(imageUrl);
  string imageData = Convert.ToBase64String(imageBytes);

  var parameters = new MessageCreateParams
  {
      Model = Model.ClaudeOpus5_5,
      MaxTokens = 1024,
      Messages =
      [
          new UserMessageParam
          {
              Content = new List<ContentBlockParam>
              {
                  new ContentBlockParam(new ImageBlockParam(
                      new ImageBlockParamSource(new Base64ImageSource()
                      {
                          Data = imageData,
                          MediaType = MediaType.ImageJpeg,
                      })
                  )),
                  new ContentBlockParam(new TextBlockParam("What is in the above image?")),
              },
          }
      ]
  };

  var message = await client.Messages.Create(parameters);
  Console.WriteLine(message);

  // Opzione 2: immagine referenziata tramite URL
  var parametersFromUrl = new MessageCreateParams
  {
      Model = Model.ClaudeOpus5_5,
      MaxTokens = 1024,
      Messages =
      [
          new UserMessageParam
          {
              Content = new List<ContentBlockParam>
              {
                  new ContentBlockParam(new ImageBlockParam(
                      new ImageBlockParamSource(new UrlImageSource()
                      {
                          Url = "https://platform.claude.com/docs/images/vision-example.jpg",
                      })
                  )),
                  new ContentBlockParam(new TextBlockParam("What is in the above image?")),
              },
          }
      ]
  };

  var messageFromUrl = await client.Messages.Create(parametersFromUrl);
  Console.WriteLine(messageFromUrl);
  ```

  ```go Go
  client := anthropic.NewClient()

  // Opzione 1: immagine codificata in base64
  imageURL := "https://platform.claude.com/docs/images/vision-example.jpg"

  req, err := http.NewRequest("GET", imageURL, nil)
  if err != nil {
  	log.Fatal(err)
  }
  req.Header.Set("User-Agent", "AnthropicDocsBot/1.0")

  resp, err := http.DefaultClient.Do(req)
  if err != nil {
  	log.Fatal(err)
  }
  defer resp.Body.Close()

  imageBytes, err := io.ReadAll(resp.Body)
  if err != nil {
  	log.Fatal(err)
  }
  imageData := base64.StdEncoding.EncodeToString(imageBytes)

  message, err := client.Messages.New(context.TODO(), anthropic.MessageNewParams{
  	Model:     anthropic.ModelClaudeOpus5_5,
  	MaxTokens: 1024,
  	Messages: []anthropic.MessageParam{
  		anthropic.NewUserMessage(
  			anthropic.NewImageBlockBase64("image/jpeg", imageData),
  			anthropic.NewTextBlock("What is in the above image?"),
  		),
  	},
  })
  if err != nil {
  	log.Fatal(err)
  }
  fmt.Println(message)

  // Opzione 2: immagine referenziata tramite URL
  messageFromURL, err := client.Messages.New(context.TODO(), anthropic.MessageNewParams{
  	Model:     anthropic.ModelClaudeOpus5_5,
  	MaxTokens: 1024,
  	Messages: []anthropic.MessageParam{
  		anthropic.NewUserMessage(
  			anthropic.NewImageBlock(anthropic.URLImageSourceParam{
  				URL: "https://platform.claude.com/docs/images/vision-example.jpg",
  			}),
  			anthropic.NewTextBlock("What is in the above image?"),
  		),
  	},
  })
  if err != nil {
  	log.Fatal(err)
  }
  fmt.Println(messageFromURL)
  ```

  ```java Java
  AnthropicClient client = AnthropicOkHttpClient.fromEnv();

  // Opzione 1: immagine codificata in base64
  String imageUrl = "https://platform.claude.com/docs/images/vision-example.jpg";

  HttpClient httpClient = HttpClient.newHttpClient();
  HttpRequest request = HttpRequest.newBuilder()
      .uri(URI.create(imageUrl))
      .build();
  HttpResponse<byte[]> response = httpClient.send(request, HttpResponse.BodyHandlers.ofByteArray());
  String imageData = Base64.getEncoder().encodeToString(response.body());

  List<ContentBlockParam> base64Content = List.of(
      ContentBlockParam.ofImage(
          ImageBlockParam.builder()
              .source(Base64ImageSource.builder()
                  .data(imageData)
                  .mediaType(Base64ImageSource.MediaType.IMAGE_JPEG)
                  .build())
              .build()),
      ContentBlockParam.ofText(
          TextBlockParam.builder()
              .text("What is in the above image?")
              .build())
  );

  Message message = client.messages().create(
      MessageCreateParams.builder()
          .model(Model.CLAUDE_OPUS_5_5)
          .maxTokens(1024L)
          .addUserMessageOfContentBlockParams(base64Content)
          .build());
  System.out.println(message);

  // Opzione 2: immagine referenziata tramite URL
  List<ContentBlockParam> urlContent = List.of(
      ContentBlockParam.ofImage(
          ImageBlockParam.builder()
              .source(UrlImageSource.builder()
                  .url("https://platform.claude.com/docs/images/vision-example.jpg")
                  .build())
              .build()),
      ContentBlockParam.ofText(
          TextBlockParam.builder()
              .text("What is in the above image?")
              .build())
  );

  Message messageFromUrl = client.messages().create(
      MessageCreateParams.builder()
          .model(Model.CLAUDE_OPUS_5_5)
          .maxTokens(1024L)
          .addUserMessageOfContentBlockParams(urlContent)
          .build());
  System.out.println(messageFromUrl);
  ```

  ```php PHP
  $client = new Client();

  // Opzione 1: immagine codificata in base64
  $image_url = 'https://platform.claude.com/docs/images/vision-example.jpg';
  $image_media_type = 'image/jpeg';
  $image_data = base64_encode(file_get_contents($image_url));

  $message = $client->messages->create(
      maxTokens: 1024,
      messages: [
          [
              'role' => 'user',
              'content' => [
                  [
                      'type' => 'image',
                      'source' => [
                          'type' => 'base64',
                          'media_type' => $image_media_type,
                          'data' => $image_data,
                      ],
                  ],
                  [
                      'type' => 'text',
                      'text' => 'What is in the above image?',
                  ],
              ],
          ],
      ],
      model: 'claude-opus-5-5',
  );
  echo $message;

  // Opzione 2: immagine referenziata tramite URL
  $message_from_url = $client->messages->create(
      maxTokens: 1024,
      messages: [
          [
              'role' => 'user',
              'content' => [
                  [
                      'type' => 'image',
                      'source' => [
                          'type' => 'url',
                          'url' => 'https://platform.claude.com/docs/images/vision-example.jpg',
                      ],
                  ],
                  [
                      'type' => 'text',
                      'text' => 'What is in the above image?',
                  ],
              ],
          ],
      ],
      model: 'claude-opus-5-5',
  );
  echo $message_from_url;
  ```

  ```ruby Ruby
  require "base64"
  require "net/http"

  client = Anthropic::Client.new

  # Opzione 1: immagine codificata in base64
  image_url = "https://platform.claude.com/docs/images/vision-example.jpg"
  image_media_type = "image/jpeg"
  image_data = Base64.strict_encode64(Net::HTTP.get(URI(image_url)))

  message = client.messages.create(
    model: "claude-opus-5-5",
    max_tokens: 1024,
    messages: [
      {
        role: "user",
        content: [
          {
            type: "image",
            source: {
              type: "base64",
              media_type: image_media_type,
              data: image_data
            }
          },
          {
            type: "text",
            text: "What is in the above image?"
          }
        ]
      }
    ]
  )
  puts message

  # Opzione 2: immagine referenziata tramite URL
  message_from_url = client.messages.create(
    model: "claude-opus-5-5",
    max_tokens: 1024,
    messages: [
      {
        role: "user",
        content: [
          {
            type: "image",
            source: {
              type: "url",
              url: "https://platform.claude.com/docs/images/vision-example.jpg"
            }
          },
          {
            type: "text",
            text: "What is in the above image?"
          }
        ]
      }
    ]
  )
  puts message_from_url
  ```
</CodeGroup>

```json Output
{
  "id": "msg_011CdKmWtV3oFx1C5yUbf5CY",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "This image is a beautiful minimalist/flat-design illustration of a sunset landscape. Here's what it contains:\n\n**Sky & Sun:**\n- A warm gradient sky transitioning from golden-yellow at the top to deep orange toward the horizon\n- A large pale yellow sun positioned in the upper-right area\n\n**Birds:**\n- Three small silhouetted birds flying in the upper-left portion of the sky, depicted as simple \"M\" or \"v\" shapes\n\n**Mountains:**\n- Multiple layered mountain peaks in purple and maroon tones\n- The mountains overlap to create depth, with varying shades of dusty purple and deep burgundy\n\n**Water:**\n- A dark purple body of water at the bottom of the image\n- A reflection of the sun shown as horizontal cream/peach colored lines in the center-bottom area\n\nThe overall style is clean, geometric, and uses a warm sunset color palette (oranges, yellows, purples, and maroons), giving it a peaceful, serene aesthetic typical of modern vector/flat design artwork."
    }
  ],
  "model": "claude-opus-5-5",
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 1030,
    "output_tokens": 350
  }
}
```

## Passaggi successivi

<CardGroup cols={2}>
  <Card title="Motivi di arresto e fallback" icon="list" href="https://platform.claude.com/docs/it/build-with-claude/handling-stop-reasons">
    Gestisci ogni valore di `stop_reason` e decidi cosa fare quando una risposta termina.
  </Card>

  <Card title="Uso degli strumenti con Claude" icon="wrench" href="https://platform.claude.com/docs/it/agents-and-tools/tool-use/overview">
    Fornisci a Claude strumenti per chiamare servizi e API esterni dall’interno dell’API Messages.
  </Card>

  <Card title="Strumento computer use" icon="computer" href="https://platform.claude.com/docs/it/agents-and-tools/tool-use/computer-use-tool">
    Controlla ambienti desktop con l’API Messages.
  </Card>

  <Card title="Strumento browser use" icon="browser" href="https://platform.claude.com/docs/it/agents-and-tools/tool-use/browser-use-tool">
    Consenti a Claude di navigare, leggere e interagire con pagine web in un browser che esegui tu.
  </Card>

  <Card title="Output strutturati" icon="code-brackets" href="https://platform.claude.com/docs/it/build-with-claude/structured-outputs">
    Ottieni da Claude output JSON garantito e convalidato tramite schema.
  </Card>

  <Card title="Budget delle attività" icon="gauge" href="https://platform.claude.com/docs/it/build-with-claude/task-budgets">
    Imposta un budget di token indicativo su un intero ciclo agentico con `output_config.task_budget`.
  </Card>
</CardGroup>

## Disponibilità

La funzionalità Messages API è disponibile su tutti i modelli e idonea per conservazione zero dei dati e preparazione HIPAA.

### Supporto API

Disponibile

### Gestione dei dati

[Panoramica sulla gestione dei dati](https://platform.claude.com/docs/it/manage-claude/api-and-data-retention#feature-eligibility)

| Elemento | Conservazione zero dei dati                                                                                                                             | HIPAA  |
| -------- | ------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ |
| Stato    | Idoneo (esclude [Modelli coperti](https://platform.claude.com/docs/it/manage-claude/api-and-data-retention#model-specific-data-retention-requirements)) | Idoneo |
| Dettagli | Chiamate API standard per generare le risposte di Claude.                                                                                               | —      |

Endpoint applicabili:

* `/v1/messages`
