Skip to main content
Glama

Genkit MCP

Official
by firebase
vertexai_live_test.go11.6 kB
// Copyright 2025 Google LLC // // Licensed under the Apache License, Version 2.0 (the "License"); // you may not use this file except in compliance with the License. // You may obtain a copy of the License at // // http://www.apache.org/licenses/LICENSE-2.0 // // Unless required by applicable law or agreed to in writing, software // distributed under the License is distributed on an "AS IS" BASIS, // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. // See the License for the specific language governing permissions and // limitations under the License. // // SPDX-License-Identifier: Apache-2.0 package googlegenai_test import ( "context" "math" "os" "strings" "testing" "github.com/firebase/genkit/go/ai" "github.com/firebase/genkit/go/genkit" "github.com/firebase/genkit/go/plugins/googlegenai" "google.golang.org/genai" ) // To run this test suite: go test -v -run TestVertexAI func TestVertexAILive(t *testing.T) { projectID, ok := requireEnv("GOOGLE_CLOUD_PROJECT") if !ok { t.Skipf("GOOGLE_CLOUD_PROJECT env var not set") } location, ok := requireEnv("GOOGLE_CLOUD_LOCATION") if !ok { t.Log("GOOGLE_CLOUD_LOCATION env var not set, defaulting to us-central1") location = "us-central1" } ctx := context.Background() g := genkit.Init(ctx, genkit.WithDefaultModel("vertexai/gemini-2.0-flash"), genkit.WithPlugins(&googlegenai.VertexAI{ProjectID: projectID, Location: location}), ) embedder := googlegenai.VertexAIEmbedder(g, "textembedding-gecko@003") gablorkenTool := genkit.DefineTool(g, "gablorken", "use this tool when the user asks to calculate a gablorken", func(ctx *ai.ToolContext, input struct { Value float64 Over float64 }, ) (float64, error) { return math.Pow(input.Value, input.Over), nil }, ) t.Run("model", func(t *testing.T) { resp, err := genkit.Generate(ctx, g, ai.WithPrompt("Which country was Napoleon the emperor of?")) if err != nil { t.Fatal(err) } out := resp.Message.Content[0].Text if !strings.Contains(out, "France") { t.Errorf("got \"%s\", expecting it would contain \"France\"", out) } if resp.Request == nil { t.Error("Request field not set properly") } if resp.Usage.InputTokens == 0 || resp.Usage.OutputTokens == 0 || resp.Usage.TotalTokens == 0 { t.Errorf("Empty usage stats %#v", *resp.Usage) } }) t.Run("streaming", func(t *testing.T) { out := "" parts := 0 final, err := genkit.Generate(ctx, g, ai.WithPrompt("Write one paragraph about the Golden State Warriors."), ai.WithStreaming(func(ctx context.Context, c *ai.ModelResponseChunk) error { parts++ for _, p := range c.Content { out += p.Text } return nil })) if err != nil { t.Fatal(err) } out2 := "" for _, p := range final.Message.Content { out2 += p.Text } if out != out2 { t.Errorf("streaming and final should contain the same text.\nstreaming:%s\nfinal:%s", out, out2) } const want = "Golden" if !strings.Contains(out, want) { t.Errorf("got %q, expecting it to contain %q", out, want) } if parts == 1 { // Check if streaming actually occurred. t.Errorf("expecting more than one part") } if final.Usage.InputTokens == 0 || final.Usage.OutputTokens == 0 || final.Usage.TotalTokens == 0 { t.Errorf("Empty usage stats %#v", *final.Usage) } }) t.Run("tool", func(t *testing.T) { resp, err := genkit.Generate(ctx, g, ai.WithPrompt("what is a gablorken of 2 over 3.5?"), ai.WithTools(gablorkenTool)) if err != nil { t.Fatal(err) } out := resp.Message.Content[0].Text if !strings.Contains(out, "12.25") { t.Errorf("got %s, expecting it to contain \"12.25\"", out) } }) t.Run("embedder", func(t *testing.T) { res, err := genkit.Embed(ctx, g, ai.WithEmbedder(embedder), ai.WithTextDocs("time flies like an arrow", "fruit flies like a banana"), ) if err != nil { t.Fatal(err) } // There's not a whole lot we can test about the result. // Just do a few sanity checks. for _, de := range res.Embeddings { out := de.Embedding if len(out) < 100 { t.Errorf("embedding vector looks too short: len(out)=%d", len(out)) } var normSquared float32 for _, x := range out { normSquared += x * x } if normSquared < 0.9 || normSquared > 1.1 { t.Errorf("embedding vector not unit length: %f", normSquared) } } }) t.Run("cache", func(t *testing.T) { if *cache == "" { t.Skip("no cache contents provided, use -cache flag") } textContent, err := os.ReadFile(*cache) if err != nil { t.Fatal(err) } resp, err := genkit.Generate(ctx, g, ai.WithMessages( ai.NewUserTextMessage(string(textContent)).WithCacheTTL(360), ), ai.WithPrompt("write a summary of the content")) if err != nil { t.Fatal(err) } // inspect metadata just to make sure the cache was created m := resp.Message.Metadata cacheName := "" if cache, ok := m["cache"].(map[string]any); ok { if n, ok := cache["name"].(string); ok { if n == "" { t.Fatal("expecting a cache name, but got nothing") } cacheName = n } else { t.Fatalf("cache name should be a string but got %T", n) } } else { t.Fatalf("cache name should be a map but got %T", cache) } resp, err = genkit.Generate(ctx, g, ai.WithMessages(resp.History()...), ai.WithPrompt("rewrite the previous summary but now talking like a pirate, say Ahoy a lot of times"), ) if err != nil { t.Fatal(err) } text := resp.Text() if !strings.Contains(text, "Ahoy") { t.Fatalf("expecting a response as a pirate but got %v", text) } // cache metadata should have not changed... if cache, ok := m["cache"].(map[string]any); ok { if n, ok := cache["name"].(string); ok { if n == "" { t.Fatal("expecting a cache name, but got nothing") } if cacheName != n { t.Fatalf("cache name mismatch, want: %s, got: %s", cacheName, n) } } else { t.Fatalf("cache name should be a string but got %T", n) } } else { t.Fatalf("cache name should be a map but got %T", cache) } }) t.Run("media content (inline data)", func(t *testing.T) { i, err := fetchImgAsBase64() if err != nil { t.Fatal(err) } resp, err := genkit.Generate(ctx, g, ai.WithSystem("You are a pirate expert in animals, your response should include the name of the animal in the provided image"), ai.WithMessages( ai.NewUserMessage( ai.NewTextPart("do you know which animal is in the image?"), ai.NewMediaPart("image/jpg", "data:image/jpg;base64,"+i), ), ), ) if err != nil { t.Fatal(err) } if !strings.Contains(strings.ToLower(resp.Text()), "cat") { t.Fatalf("image detection failed, want: cat, got: %s", resp.Text()) } }) t.Run("media content", func(t *testing.T) { resp, err := genkit.Generate(ctx, g, ai.WithMessages( ai.NewUserMessage( ai.NewTextPart("do you know what's the video about?"), ai.NewMediaPart("video/mp4", `https://www.youtube.com/watch?v=_6FYhqGgel8`), ), ), ) if err != nil { t.Fatal(err) } if !strings.Contains(resp.Text(), "Mario Kart") { t.Fatalf("image detection failed, want: Mario Kart, got: %s", resp.Text()) } }) t.Run("data content (inline data)", func(t *testing.T) { i, err := fetchImgAsBase64() if err != nil { t.Fatal(err) } resp, err := genkit.Generate(ctx, g, ai.WithSystem("You are a pirate expert in animals, your response should include the name of the animal in the image provided"), ai.WithMessages( ai.NewUserMessage( ai.NewTextPart("do you know which animal is in the image?"), ai.NewDataPart("data:image/jpg;base64,"+i), ), ), ) if err != nil { t.Fatal(err) } if !strings.Contains(strings.ToLower(resp.Text()), "cat") { t.Fatalf("image detection failed, want: cat, got: %s", resp.Text()) } }) t.Run("image generation", func(t *testing.T) { if location != "global" { t.Skipf("image generation in Vertex AI is only supported in region: global, got: %s", location) } m := googlegenai.VertexAIModel(g, "gemini-2.0-flash-preview-image-generation") resp, err := genkit.Generate(ctx, g, ai.WithConfig(genai.GenerateContentConfig{ ResponseModalities: []string{"IMAGE", "TEXT"}, }), ai.WithMessages( ai.NewUserTextMessage("generate an image of a dog wearing a black tejana while playing the accordion"), ), ai.WithModel(m), ) if err != nil { t.Fatal(err) } if len(resp.Message.Content) == 0 { t.Fatal("empty response") } foundMediaPart := false for _, part := range resp.Message.Content { if part.ContentType == "image/png" { foundMediaPart = true if part.Kind != ai.PartMedia { t.Errorf("expecting part to be Media type but got: %q", part.Kind) } if part.Text == "" { t.Error("empty response") } } } if !foundMediaPart { t.Error("no media found in the response message") } }) t.Run("constrained generation", func(t *testing.T) { type outFormat struct { Country string } resp, err := genkit.Generate(ctx, g, ai.WithPrompt("Which country was Napoleon the emperor of?"), ai.WithOutputType(outFormat{}), ) if err != nil { t.Fatal(err) } var ans outFormat err = resp.Output(&ans) if err != nil { t.Fatal(err) } const want = "France" if ans.Country != want { t.Errorf("got %q, expecting %q", ans.Country, want) } if resp.Request == nil { t.Error("Request field not set properly") } if resp.Usage.InputTokens == 0 || resp.Usage.OutputTokens == 0 || resp.Usage.TotalTokens == 0 { t.Errorf("Empty usage stats %#v", *resp.Usage) } }) t.Run("thinking enabled", func(t *testing.T) { if location != "global" && location != "us-central1" { t.Skipf("thinking in Vertex AI is only supported in these regions: [global, us-central1], got: %q", location) } m := googlegenai.VertexAIModel(g, "gemini-2.5-flash-preview-05-20") resp, err := genkit.Generate(ctx, g, ai.WithConfig( genai.GenerateContentConfig{ Temperature: genai.Ptr[float32](1), ThinkingConfig: &genai.ThinkingConfig{ IncludeThoughts: true, ThinkingBudget: genai.Ptr[int32](1024), }, }, ), ai.WithPrompt(`how is a black hole born?`), ai.WithModel(m), ) if err != nil { t.Fatal(err) } if resp.Reasoning() == "" { t.Error("expected reasoning contents but got empty") } if resp.Text() == "" { t.Error("expecting response output, got empty") } if resp.Usage.ThoughtsTokens == 0 { t.Error("expecting thought token count, got 0") } }) t.Run("thinking disabled", func(t *testing.T) { if location != "global" && location != "us-central1" { t.Skipf("thinking in Vertex AI is only supported in these regions: [global, us-central1], got: %q", location) } m := googlegenai.VertexAIModel(g, "gemini-2.5-flash-preview-05-20") resp, err := genkit.Generate(ctx, g, ai.WithConfig( genai.GenerateContentConfig{ Temperature: genai.Ptr[float32](1), ThinkingConfig: &genai.ThinkingConfig{ IncludeThoughts: false, ThinkingBudget: genai.Ptr[int32](0), }, }, ), ai.WithPrompt(`how is a black hole born?`), ai.WithModel(m), ) if err != nil { t.Fatal(err) } if resp.Reasoning() != "" { t.Error("expected reasoning contents but got content") } if resp.Text() == "" { t.Error("expecting response output, got empty") } if resp.Usage.ThoughtsTokens > 0 { t.Errorf("expecting 0 thought tokens, got %d", resp.Usage.ThoughtsTokens) } }) }

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/firebase/genkit'

If you have feedback or need assistance with the MCP directory API, please join our Discord server