co-mono/packages/ai/src/providers/google.ts
Mario Zechner 2cfd8ff3c3 fix(ai): Use API type instead of model for message compatibility checks
- Add getApi() method to all providers to identify the API type
- Add api field to AssistantMessage to track which API generated it
- Update transformMessages to check API compatibility instead of model
- Fixes issue where OpenAI Responses API failed when switching models
- Preserves thinking blocks and signatures when staying within same API
2025-09-02 00:20:06 +02:00

391 lines
10 KiB
TypeScript

import {
type Content,
type FinishReason,
FunctionCallingConfigMode,
type GenerateContentConfig,
type GenerateContentParameters,
GoogleGenAI,
type Part,
setDefaultBaseUrls,
} from "@google/genai";
import { calculateCost } from "../models.js";
import type {
AssistantMessage,
Context,
LLM,
LLMOptions,
Message,
Model,
StopReason,
TextContent,
ThinkingContent,
Tool,
ToolCall,
} from "../types.js";
import { transformMessages } from "./utils.js";
export interface GoogleLLMOptions extends LLMOptions {
toolChoice?: "auto" | "none" | "any";
thinking?: {
enabled: boolean;
budgetTokens?: number; // -1 for dynamic, 0 to disable
};
}
export class GoogleLLM implements LLM<GoogleLLMOptions> {
private client: GoogleGenAI;
private modelInfo: Model;
constructor(model: Model, apiKey?: string) {
if (!apiKey) {
if (!process.env.GEMINI_API_KEY) {
throw new Error(
"Gemini API key is required. Set GEMINI_API_KEY environment variable or pass it as an argument.",
);
}
apiKey = process.env.GEMINI_API_KEY;
}
this.client = new GoogleGenAI({ apiKey });
this.modelInfo = model;
}
getModel(): Model {
return this.modelInfo;
}
getApi(): string {
return "google-generative-ai";
}
async generate(context: Context, options?: GoogleLLMOptions): Promise<AssistantMessage> {
const output: AssistantMessage = {
role: "assistant",
content: [],
api: this.getApi(),
provider: this.modelInfo.provider,
model: this.modelInfo.id,
usage: {
input: 0,
output: 0,
cacheRead: 0,
cacheWrite: 0,
cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0, total: 0 },
},
stopReason: "stop",
};
try {
const contents = this.convertMessages(context.messages);
// Build generation config
const generationConfig: GenerateContentConfig = {};
if (options?.temperature !== undefined) {
generationConfig.temperature = options.temperature;
}
if (options?.maxTokens !== undefined) {
generationConfig.maxOutputTokens = options.maxTokens;
}
// Build the config object
const config: GenerateContentConfig = {
...(Object.keys(generationConfig).length > 0 && generationConfig),
...(context.systemPrompt && { systemInstruction: context.systemPrompt }),
...(context.tools && { tools: this.convertTools(context.tools) }),
};
// Add tool config if needed
if (context.tools && options?.toolChoice) {
config.toolConfig = {
functionCallingConfig: {
mode: this.mapToolChoice(options.toolChoice),
},
};
}
// Add thinking config if enabled and model supports it
if (options?.thinking?.enabled && this.modelInfo.reasoning) {
config.thinkingConfig = {
includeThoughts: true,
...(options.thinking.budgetTokens !== undefined && { thinkingBudget: options.thinking.budgetTokens }),
};
}
// Abort signal
if (options?.signal) {
if (options.signal.aborted) {
throw new Error("Request aborted");
}
config.abortSignal = options.signal;
}
// Build the request parameters
const params: GenerateContentParameters = {
model: this.modelInfo.id,
contents,
config,
};
const stream = await this.client.models.generateContentStream(params);
options?.onEvent?.({ type: "start", model: this.modelInfo.id, provider: this.modelInfo.provider });
let currentBlock: TextContent | ThinkingContent | null = null;
for await (const chunk of stream) {
// Extract parts from the chunk
const candidate = chunk.candidates?.[0];
if (candidate?.content?.parts) {
for (const part of candidate.content.parts) {
if (part.text !== undefined) {
const isThinking = part.thought === true;
// Check if we need to switch blocks
if (
!currentBlock ||
(isThinking && currentBlock.type !== "thinking") ||
(!isThinking && currentBlock.type !== "text")
) {
if (currentBlock) {
if (currentBlock.type === "text") {
options?.onEvent?.({ type: "text_end", content: currentBlock.text });
} else {
options?.onEvent?.({ type: "thinking_end", content: currentBlock.thinking });
}
}
// Start new block
if (isThinking) {
currentBlock = { type: "thinking", thinking: "", thinkingSignature: undefined };
options?.onEvent?.({ type: "thinking_start" });
} else {
currentBlock = { type: "text", text: "" };
options?.onEvent?.({ type: "text_start" });
}
output.content.push(currentBlock);
}
// Append content to current block
if (currentBlock.type === "thinking") {
currentBlock.thinking += part.text;
currentBlock.thinkingSignature = part.thoughtSignature;
options?.onEvent?.({
type: "thinking_delta",
content: currentBlock.thinking,
delta: part.text,
});
} else {
currentBlock.text += part.text;
options?.onEvent?.({ type: "text_delta", content: currentBlock.text, delta: part.text });
}
}
// Handle function calls
if (part.functionCall) {
if (currentBlock) {
if (currentBlock.type === "text") {
options?.onEvent?.({ type: "text_end", content: currentBlock.text });
} else {
options?.onEvent?.({ type: "thinking_end", content: currentBlock.thinking });
}
currentBlock = null;
}
// Add tool call
const toolCallId = part.functionCall.id || `${part.functionCall.name}_${Date.now()}`;
const toolCall: ToolCall = {
type: "toolCall",
id: toolCallId,
name: part.functionCall.name || "",
arguments: part.functionCall.args as Record<string, any>,
};
output.content.push(toolCall);
options?.onEvent?.({ type: "toolCall", toolCall });
}
}
}
// Map finish reason
if (candidate?.finishReason) {
output.stopReason = this.mapStopReason(candidate.finishReason);
// Check if we have tool calls in blocks
if (output.content.some((b) => b.type === "toolCall")) {
output.stopReason = "toolUse";
}
}
// Capture usage metadata if available
if (chunk.usageMetadata) {
output.usage = {
input: chunk.usageMetadata.promptTokenCount || 0,
output:
(chunk.usageMetadata.candidatesTokenCount || 0) + (chunk.usageMetadata.thoughtsTokenCount || 0),
cacheRead: chunk.usageMetadata.cachedContentTokenCount || 0,
cacheWrite: 0,
cost: {
input: 0,
output: 0,
cacheRead: 0,
cacheWrite: 0,
total: 0,
},
};
calculateCost(this.modelInfo, output.usage);
}
}
// Finalize last block
if (currentBlock) {
if (currentBlock.type === "text") {
options?.onEvent?.({ type: "text_end", content: currentBlock.text });
} else {
options?.onEvent?.({ type: "thinking_end", content: currentBlock.thinking });
}
}
options?.onEvent?.({ type: "done", reason: output.stopReason, message: output });
return output;
} catch (error) {
output.stopReason = "error";
output.error = error instanceof Error ? error.message : JSON.stringify(error);
options?.onEvent?.({ type: "error", error: output.error });
return output;
}
}
private convertMessages(messages: Message[]): Content[] {
const contents: Content[] = [];
// Transform messages for cross-provider compatibility
const transformedMessages = transformMessages(messages, this.modelInfo, this.getApi());
for (const msg of transformedMessages) {
if (msg.role === "user") {
// Handle both string and array content
if (typeof msg.content === "string") {
contents.push({
role: "user",
parts: [{ text: msg.content }],
});
} else {
// Convert array content to Google format
const parts: Part[] = msg.content.map((item) => {
if (item.type === "text") {
return { text: item.text };
} else {
// Image content - Google uses inlineData
return {
inlineData: {
mimeType: item.mimeType,
data: item.data,
},
};
}
});
const filteredParts = !this.modelInfo?.input.includes("image")
? parts.filter((p) => p.text !== undefined)
: parts;
contents.push({
role: "user",
parts: filteredParts,
});
}
} else if (msg.role === "assistant") {
const parts: Part[] = [];
// Process content blocks
for (const block of msg.content) {
if (block.type === "text") {
parts.push({ text: block.text });
} else if (block.type === "thinking") {
const thinkingPart: Part = {
thought: true,
thoughtSignature: block.thinkingSignature,
text: block.thinking,
};
parts.push(thinkingPart);
} else if (block.type === "toolCall") {
parts.push({
functionCall: {
id: block.id,
name: block.name,
args: block.arguments,
},
});
}
}
if (parts.length > 0) {
contents.push({
role: "model",
parts,
});
}
} else if (msg.role === "toolResult") {
contents.push({
role: "user",
parts: [
{
functionResponse: {
id: msg.toolCallId,
name: msg.toolName,
response: {
result: msg.content,
isError: msg.isError,
},
},
},
],
});
}
}
return contents;
}
private convertTools(tools: Tool[]): any[] {
return [
{
functionDeclarations: tools.map((tool) => ({
name: tool.name,
description: tool.description,
parameters: tool.parameters,
})),
},
];
}
private mapToolChoice(choice: string): FunctionCallingConfigMode {
switch (choice) {
case "auto":
return FunctionCallingConfigMode.AUTO;
case "none":
return FunctionCallingConfigMode.NONE;
case "any":
return FunctionCallingConfigMode.ANY;
default:
return FunctionCallingConfigMode.AUTO;
}
}
private mapStopReason(reason: FinishReason): StopReason {
switch (reason) {
case "STOP":
return "stop";
case "MAX_TOKENS":
return "length";
case "BLOCKLIST":
case "PROHIBITED_CONTENT":
case "SPII":
case "SAFETY":
case "IMAGE_SAFETY":
return "safety";
case "RECITATION":
return "safety";
case "FINISH_REASON_UNSPECIFIED":
case "OTHER":
case "LANGUAGE":
case "MALFORMED_FUNCTION_CALL":
case "UNEXPECTED_TOOL_CALL":
return "error";
default:
return "stop";
}
}
}