Add speech and video Inference APIs with multi-language examples

specification/DigitalOcean-public.v2.yaml

@@ -3981,6 +3981,23 @@ paths:
     post:
       $ref: 'resources/inference/inference_async_invoke.yml'
 
+
+  /v1/audio/speech:
+    post:
+      $ref: 'resources/inference/inference_create_speech.yml'
+
+  /v1/videos:
+    post:
+      $ref: 'resources/inference/inference_create_video.yml'
+
+  /v1/videos/{video_id}:
+    get:
+      $ref: 'resources/inference/inference_get_video.yml'
+
+  /v1/videos/{video_id}/content:
+    get:
+      $ref: 'resources/inference/inference_get_video_content.yml'
+
   /v1/batches/files:
     post:
       $ref: 'resources/inference/inference_create_batch_file.yml'

specification/resources/inference/examples/curl/inference_create_speech.yml

@@ -0,0 +1,13 @@
+lang: cURL
+source: |-
+  curl -sS https://inference.do-ai.run/v1/audio/speech \
+    -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
+    -H "Content-Type: application/json" \
+    -d '{
+      "model": "qwen3-tts-voicedesign",
+      "input": "Hello, world!",
+      "voice": "alloy",
+      "response_format": "mp3",
+      "instructions": "Speak naturally."
+    }' \
+    --output output.mp3

specification/resources/inference/examples/curl/inference_create_video.yml

@@ -0,0 +1,11 @@
+lang: cURL
+source: |-
+  curl -X POST https://inference.do-ai.run/v1/videos \
+    -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
+    -H "Content-Type: application/json" \
+    -d '{
+      "model": "wan2-2-t2v-a14b",
+      "prompt": "Two anthropomorphic cats in boxing gear fight intensely on a spotlighted stage",
+      "size": "1280x720",
+      "fps": 16
+    }'

specification/resources/inference/examples/curl/inference_get_video.yml

@@ -0,0 +1,4 @@
+lang: cURL
+source: |-
+  curl https://inference.do-ai.run/v1/videos/$VIDEO_ID \
+    -H "Authorization: Bearer $DIGITALOCEAN_TOKEN"

specification/resources/inference/examples/curl/inference_get_video_content.yml

@@ -0,0 +1,5 @@
+lang: cURL
+source: |-
+  curl https://inference.do-ai.run/v1/videos/$VIDEO_ID/content \
+    -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
+    --output output.mp4

specification/resources/inference/examples/dots/inference_create_speech.yml

@@ -0,0 +1,19 @@
+lang: JavaScript
+source: |-
+  import { writeFileSync } from "fs";
+  import { InferenceClient } from "@digitalocean/dots";
+
+  const client = new InferenceClient({
+      apiKey: process.env.DIGITALOCEAN_TOKEN,
+  });
+
+  const response = await client.audio.speech.create({
+      model: "qwen3-tts-voicedesign",
+      input: "Hello, world!",
+      voice: "alloy",
+      response_format: "mp3",
+      instructions: "Speak naturally.",
+  });
+
+  const audioBytes = Buffer.from(response.data[0].b64_json, "base64");
+  writeFileSync("output.mp3", audioBytes);

specification/resources/inference/examples/dots/inference_create_video.yml

@@ -0,0 +1,16 @@
+lang: JavaScript
+source: |-
+  import { InferenceClient } from "@digitalocean/dots";
+
+  const client = new InferenceClient({
+      apiKey: process.env.DIGITALOCEAN_TOKEN,
+  });
+
+  const job = await client.video.generations.create({
+      model: "wan2-2-t2v-a14b",
+      prompt: "Two anthropomorphic cats in boxing gear fight intensely on a spotlighted stage",
+      size: "1280x720",
+      fps: 16,
+  });
+
+  console.log(job.id, job.status);

specification/resources/inference/examples/dots/inference_get_video.yml

@@ -0,0 +1,10 @@
+lang: JavaScript
+source: |-
+  import { InferenceClient } from "@digitalocean/dots";
+
+  const client = new InferenceClient({
+      apiKey: process.env.DIGITALOCEAN_TOKEN,
+  });
+
+  const job = await client.video.generations.retrieve(process.env.VIDEO_ID);
+  console.log(job.id, job.status);

specification/resources/inference/examples/dots/inference_get_video_content.yml

@@ -0,0 +1,11 @@
+lang: JavaScript
+source: |-
+  import { writeFileSync } from "fs";
+  import { InferenceClient } from "@digitalocean/dots";
+
+  const client = new InferenceClient({
+      apiKey: process.env.DIGITALOCEAN_TOKEN,
+  });
+
+  const content = await client.video.generations.content(process.env.VIDEO_ID);
+  writeFileSync("output.mp4", Buffer.from(await content.arrayBuffer()));

specification/resources/inference/examples/go/agentInference_create_chat_completion.yml

@@ -32,6 +32,6 @@ source: |-
       }
 
       if msg := completion.Choices[0].Message.Content; msg != nil {
-          fmt.Println(*msg)
+          fmt.Println(msg.Text())
       }
   }

specification/resources/inference/examples/go/inference_create_chat_completion.yml

@@ -25,6 +25,6 @@ source: |-
       }
 
       if msg := completion.Choices[0].Message.Content; msg != nil {
-          fmt.Println(*msg)
+          fmt.Println(msg.Text())
       }
   }

specification/resources/inference/examples/go/inference_create_chat_completion_kb.yml

@@ -0,0 +1,37 @@
+lang: Go
+label: Knowledge base
+source: |-
+  package main
+
+  import (
+      "context"
+      "encoding/json"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      completion, _, err := client.Chat.Completions.New(ctx, &godo.ChatCompletionNewParams{
+          Model: "openai-gpt-oss-20b",
+          Messages: []godo.ChatCompletionMessage{
+              godo.UserMessage("What are some features of DigitalOcean Inference?"),
+          },
+          Tools: []godo.ChatCompletionTool{
+              godo.ChatKnowledgeBaseRetrievalTool(os.Getenv("KNOWLEDGE_BASE_ID")),
+          },
+          ToolChoice: json.RawMessage(`"auto"`),
+          MaxTokens:  godo.PtrTo(1024),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      if msg := completion.Choices[0].Message.Content; msg != nil {
+          fmt.Println(msg.Text())
+      }
+  }

specification/resources/inference/examples/go/inference_create_chat_completion_mcp.yml

@@ -0,0 +1,42 @@
+lang: Go
+label: MCP tool
+source: |-
+  package main
+
+  import (
+      "context"
+      "encoding/json"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      completion, _, err := client.Chat.Completions.New(ctx, &godo.ChatCompletionNewParams{
+          Model: "openai-gpt-oss-20b",
+          Messages: []godo.ChatCompletionMessage{
+              godo.UserMessage("Fetch my DigitalOcean account information and summarize it in 2 bullets."),
+          },
+          Tools: []godo.ChatCompletionTool{
+              godo.ChatMCPTool(godo.MCPToolParams{
+                  ServerLabel:   "digitalocean",
+                  ServerURL:     "https://accounts.mcp.digitalocean.com/mcp",
+                  Authorization: "Bearer " + os.Getenv("DIGITALOCEAN_API_TOKEN"),
+                  AllowedTools:  []string{"account-get-information"},
+              }),
+          },
+          ToolChoice: json.RawMessage(`"required"`),
+          MaxTokens:  godo.PtrTo(512),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      if msg := completion.Choices[0].Message.Content; msg != nil {
+          fmt.Println(msg.Text())
+      }
+  }

specification/resources/inference/examples/go/inference_create_chat_completion_vlm.yml

@@ -0,0 +1,41 @@
+lang: Go
+label: Vision (VLM)
+source: |-
+  package main
+
+  import (
+      "context"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      // IMAGE_URL may be an HTTPS URL or a data:image/...;base64,... URI.
+      imageURL := os.Getenv("IMAGE_URL")
+      if imageURL == "" {
+          imageURL = "https://upload.wikimedia.org/wikipedia/commons/thumb/3/3b/Online_Bicycles.jpg/640px-Online_Bicycles.jpg"
+      }
+
+      completion, _, err := client.Chat.Completions.New(ctx, &godo.ChatCompletionNewParams{
+          Model: "nemotron-nano-12b-v2-vl",
+          Messages: []godo.ChatCompletionMessage{
+              godo.UserMessageParts(
+                  godo.TextPart("Describe this image."),
+                  godo.ImageURLPart(imageURL),
+              ),
+          },
+          MaxTokens: godo.PtrTo(512),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      if msg := completion.Choices[0].Message.Content; msg != nil {
+          fmt.Println(msg.Text())
+      }
+  }

specification/resources/inference/examples/go/inference_create_response_model_synthesis.yml

@@ -0,0 +1,33 @@
+lang: Go
+label: Model synthesis
+source: |-
+  package main
+
+  import (
+      "context"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      resp, _, err := client.Responses.New(ctx, &godo.ResponseNewParams{
+          Model: "openai-gpt-oss-20b",
+          Input: "Compare approaches for reducing inference latency.",
+          Tools: []godo.ResponseTool{
+              godo.ModelSynthesisTool(godo.ModelSynthesisToolParams{
+                  Preset: "balanced",
+              }),
+          },
+          MaxOutputTokens: godo.PtrTo(1024),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      fmt.Println(resp.OutputText())
+  }

specification/resources/inference/examples/go/inference_create_response_tool_search.yml

@@ -0,0 +1,38 @@
+lang: Go
+label: Tool search
+source: |-
+  package main
+
+  import (
+      "context"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      resp, _, err := client.Responses.New(ctx, &godo.ResponseNewParams{
+          Model: "openai-gpt-5.5",
+          Input: "Compare the current weather in zip code 94107 and 10001.",
+          Tools: []godo.ResponseTool{
+              godo.FunctionTool("get_weather", "Get weather for a zip code", map[string]interface{}{
+                  "type": "object",
+                  "properties": map[string]interface{}{
+                      "zip": map[string]interface{}{"type": "string"},
+                  },
+                  "required": []string{"zip"},
+              }),
+              godo.ToolSearchTool(),
+          },
+          MaxOutputTokens: godo.PtrTo(1024),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      fmt.Println(resp.OutputText())
+  }

specification/resources/inference/examples/go/inference_create_response_web_fetch.yml

@@ -0,0 +1,31 @@
+lang: Go
+label: Web fetch
+source: |-
+  package main
+
+  import (
+      "context"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      resp, _, err := client.Responses.New(ctx, &godo.ResponseNewParams{
+          Model: "openai-gpt-oss-20b",
+          Input: "Summarize https://www.digitalocean.com/pricing",
+          Tools: []godo.ResponseTool{
+              godo.WebFetchTool(godo.PtrTo(3)),
+          },
+          MaxOutputTokens: godo.PtrTo(1024),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      fmt.Println(resp.OutputText())
+  }

specification/resources/inference/examples/go/inference_create_response_web_search.yml

@@ -0,0 +1,31 @@
+lang: Go
+label: Web search
+source: |-
+  package main
+
+  import (
+      "context"
+      "fmt"
+      "os"
+
+      "github.com/digitalocean/godo"
+  )
+
+  func main() {
+      client := godo.NewFromToken(os.Getenv("DIGITALOCEAN_TOKEN"))
+      ctx := context.Background()
+
+      resp, _, err := client.Responses.New(ctx, &godo.ResponseNewParams{
+          Model: "openai-gpt-oss-20b",
+          Input: "What are the latest pricing changes for DigitalOcean Droplets?",
+          Tools: []godo.ResponseTool{
+              godo.WebSearchTool(godo.PtrTo(3), godo.PtrTo(5)),
+          },
+          MaxOutputTokens: godo.PtrTo(1024),
+      })
+      if err != nil {
+          panic(err)
+      }
+
+      fmt.Println(resp.OutputText())
+  }