@cartanova/qgrid-ai-sdk 2.4.1 → 2.4.3

package/README.ko.md

@@ -160,7 +160,7 @@ const { text } = await generateText({
 | 옵션 | 값 | 적용 범위 | 설명 |
 |---|---|---|---|
 | `sessionKey` | `string` | OpenAI 전용 | 멀티턴 대화 식별자. 같은 key는 같은 codex thread로 라우팅되어 prompt cache 적중 ([아래](#멀티턴-prompt-cache-sessionkey) 참조) |
-| `effort` | `"none"` \| `"minimal"` \| `"low"` \| `"medium"` \| `"high"` \| `"xhigh"` | 공통 | reasoning 모델의 추론 깊이. 기본값은 config의 `defaultEffort` (`"low"`) |
+| `effort` | `"none"` \| `"minimal"` \| `"low"` \| `"medium"` \| `"high"` \| `"xhigh"` \| `"max"` | 공통 (지원 값은 모델별 상이, 예: `"max"`는 GPT-5.6+) | reasoning 모델의 추론 깊이. 기본값은 config의 `defaultEffort` (`"low"`) |
 | `verbosity` | `"low"` \| `"medium"` \| `"high"` | OpenAI 전용 | 응답 텍스트의 상세도 |
 | `reasoningSummary` | `"auto"` \| `"concise"` \| `"detailed"` \| `"none"` | OpenAI 전용 | 추론 요약 출력 방식 |
 | `serviceTier` | `string` | OpenAI 전용 | OpenAI/codex service tier |
@@ -269,6 +269,9 @@ createQgridLogger({
 ```typescript
 type QgridSupportedModel =
   // OpenAI (based on codex app-server)
+  | "openai/gpt-5.6-sol"
+  | "openai/gpt-5.6-terra"
+  | "openai/gpt-5.6-luna"
   | "openai/gpt-5.5"
   | "openai/gpt-5.4"
   | "openai/gpt-5.2"
@@ -290,6 +293,16 @@ type QgridSupportedModel =
   | "anthropic/claude-opus-4-8"
 ```
 
+### GPT-5.6 사양
+
+| 모델 | Context (qgrid/codex 런타임) | 최대 출력 | 1M tokens당 input / cached input / output |
+|---|---:|---:|---:|
+| `openai/gpt-5.6-sol` | 372K | 128K | $5 / $0.50 / $30 |
+| `openai/gpt-5.6-terra` | 372K | 128K | $2.50 / $0.25 / $15 |
+| `openai/gpt-5.6-luna` | 372K | 128K | $1 / $0.10 / $6 |
+
+세 모델 모두 `max` reasoning effort를 지원합니다. OpenAI native API 사양은 1.05M context / 128K 최대 출력이지만, qgrid는 codex app-server 구독 경로로 실행되어 세 모델 모두 context window가 372K(95% effective, 실사용 입력 약 353K)로 보고되며 설정으로 올릴 수 없습니다. 입력이 272K tokens를 넘으면 요청 전체에 input 2x, output 1.5x 장문 컨텍스트 할증이 적용되며, cache write는 uncached input 단가의 1.25x입니다.
+
 ## 설정
 
 ```typescript

package/README.md

@@ -160,7 +160,7 @@ const { text } = await generateText({
 | Option | Values | Applies to | Description |
 |---|---|---|---|
 | `sessionKey` | `string` | OpenAI only | Multi-turn conversation identifier. The same key routes to the same codex thread for prompt-cache hits (see [below](#multi-turn-prompt-cache-sessionkey)) |
-| `effort` | `"none"` \| `"minimal"` \| `"low"` \| `"medium"` \| `"high"` \| `"xhigh"` | both providers | Reasoning depth. Defaults to the config's `defaultEffort` (`"low"`) |
+| `effort` | `"none"` \| `"minimal"` \| `"low"` \| `"medium"` \| `"high"` \| `"xhigh"` \| `"max"` | both providers (supported values are model-dependent, e.g. `"max"` is GPT-5.6+) | Reasoning depth. Defaults to the config's `defaultEffort` (`"low"`) |
 | `verbosity` | `"low"` \| `"medium"` \| `"high"` | OpenAI only | Response text verbosity |
 | `reasoningSummary` | `"auto"` \| `"concise"` \| `"detailed"` \| `"none"` | OpenAI only | Reasoning summary output mode |
 | `serviceTier` | `string` | OpenAI only | OpenAI/codex service tier |
@@ -269,6 +269,9 @@ The `qgrid()` provider has its own lifecycle, so the logger automatically suppre
 ```typescript
 type QgridSupportedModel =
   // OpenAI (based on codex app-server)
+  | "openai/gpt-5.6-sol"
+  | "openai/gpt-5.6-terra"
+  | "openai/gpt-5.6-luna"
   | "openai/gpt-5.5"
   | "openai/gpt-5.4"
   | "openai/gpt-5.2"
@@ -290,6 +293,16 @@ type QgridSupportedModel =
   | "anthropic/claude-opus-4-8"
 ```
 
+### GPT-5.6 specifications
+
+| Model | Context (qgrid/codex runtime) | Max output | Input / cached input / output per 1M tokens |
+|---|---:|---:|---:|
+| `openai/gpt-5.6-sol` | 372K | 128K | $5 / $0.50 / $30 |
+| `openai/gpt-5.6-terra` | 372K | 128K | $2.50 / $0.25 / $15 |
+| `openai/gpt-5.6-luna` | 372K | 128K | $1 / $0.10 / $6 |
+
+All GPT-5.6 models support reasoning through `max`. The OpenAI native API spec is a 1.05M context window with 128K max output, but qgrid runs on the codex app-server subscription path, where all three models report a 372K context window (95% effective — about 353K of usable input) that cannot be raised by configuration. Prompts over 272K input tokens apply a 2x input and 1.5x output surcharge to the full request; cache writes cost 1.25x the uncached input rate.
+
 ## Configuration
 
 ```typescript

package/dist/index.d.ts

@@ -20,7 +20,7 @@ type QgridProviderOptions = {
    */
   sessionKey?: string;
   /** reasoning 모델의 추론 깊이. 기본값은 qgrid config의 defaultEffort. */
-  effort?: "none" | "minimal" | "low" | "medium" | "high" | "xhigh";
+  effort?: "none" | "minimal" | "low" | "medium" | "high" | "xhigh" | "max";
   /** 응답 텍스트의 상세도. OpenAI/codex route에서만 적용된다. */
   verbosity?: "low" | "medium" | "high";
   /** reasoning 모델의 추론 요약 출력 방식. OpenAI/codex route에서만 적용된다. */
@@ -50,7 +50,7 @@ type QgridProviderOptions = {
  * provider 내부 threadCoordStore 가 sessionKey 별로 관리
  */
 
-type QgridSupportedModel = "openai/gpt-5.5" | "openai/gpt-5.4" | "openai/gpt-5.2" | "openai/gpt-5.4-mini" | "openai/gpt-5.3-codex" | "openai/gpt-5.3-codex-spark" | "anthropic/claude-haiku-4-5" | "anthropic/claude-sonnet-4" | "anthropic/claude-sonnet-4-5" | "anthropic/claude-sonnet-4-6" | "anthropic/claude-sonnet-4-7" | "anthropic/claude-sonnet-5" | "anthropic/claude-opus-4" | "anthropic/claude-opus-4-1" | "anthropic/claude-opus-4-5" | "anthropic/claude-opus-4-6" | "anthropic/claude-opus-4-7" | "anthropic/claude-opus-4-8";
+type QgridSupportedModel = "openai/gpt-5.6-sol" | "openai/gpt-5.6-terra" | "openai/gpt-5.6-luna" | "openai/gpt-5.5" | "openai/gpt-5.4" | "openai/gpt-5.2" | "openai/gpt-5.4-mini" | "openai/gpt-5.3-codex" | "openai/gpt-5.3-codex-spark" | "anthropic/claude-haiku-4-5" | "anthropic/claude-sonnet-4" | "anthropic/claude-sonnet-4-5" | "anthropic/claude-sonnet-4-6" | "anthropic/claude-sonnet-4-7" | "anthropic/claude-sonnet-5" | "anthropic/claude-opus-4" | "anthropic/claude-opus-4-1" | "anthropic/claude-opus-4-5" | "anthropic/claude-opus-4-6" | "anthropic/claude-opus-4-7" | "anthropic/claude-opus-4-8";
 type QgridLoggerConfig = {
   /** qgrid 서버 주소. default: process.env.QGRID_URL, 없으면 기본값은 "http://localhost:44900" */
   serverUrl?: string;

package/dist/index.js

@@ -3,7 +3,7 @@ const QGRID_PROVIDER_NAME = "qgrid";
 const THREAD_COORD_TTL_MS = 10 * 6e4;
 const TEXT_STREAM_ID_PREFIX = "text";
 const TOP_LEVEL_SCHEMA_WARNING = "OpenAI structured output requires \"object\". Falling back to client-side parsing.";
-const PENDING_TOOL_RESULTS_WARNING = "[qgrid] pending tool results not found in prompt, clearing client run state";
+const PENDING_TOOL_RESULTS_WARNING = "[qgrid] tool results did not match any pending qgrid run; sending request without runContext";
 //#endregion
 //#region src/utils.ts
 async function createRun(serverUrl, body) {
@@ -602,16 +602,16 @@ function createQgridLogger(config = {}) {
 //#endregion
 //#region src/index.ts
 const threadCoordStore = /* @__PURE__ */ new Map();
-const MAX_IMAGE_INPUT_DATA_URL_CHARS = 9e5;
-const MAX_IMAGE_INPUT_DATA_URL_CHARS_LABEL = `${MAX_IMAGE_INPUT_DATA_URL_CHARS / 1e3}k`;
+const MAX_IMAGE_INPUT_DATA_URL_CHARS = 9e6;
+const MAX_IMAGE_INPUT_DATA_URL_CHARS_LABEL = "9M";
 function assertImageInputFitsJsonTransport(imageUrls) {
 	let totalDataUrlChars = 0;
 	for (const url of imageUrls) {
 		if (!url.toLowerCase().startsWith("data:")) continue;
 		totalDataUrlChars += url.length;
-		if (url.length > MAX_IMAGE_INPUT_DATA_URL_CHARS) throw new Error(`qgrid: reference image input is too large for JSON transport (${url.length} chars). Compress or resize it before passing it to imageGeneration, for example as WebP/JPEG under ${MAX_IMAGE_INPUT_DATA_URL_CHARS_LABEL} total base64 data-url chars.`);
+		if (url.length > MAX_IMAGE_INPUT_DATA_URL_CHARS) throw new Error(`qgrid: image input is too large for JSON transport (${url.length} chars). Compress or resize it before passing it to imageGeneration, for example as WebP/JPEG under ${MAX_IMAGE_INPUT_DATA_URL_CHARS_LABEL} total base64 data-url chars.`);
 	}
-	if (totalDataUrlChars > MAX_IMAGE_INPUT_DATA_URL_CHARS) throw new Error(`qgrid: reference image inputs are too large for JSON transport (${totalDataUrlChars} total chars). Compress or resize them before passing them to imageGeneration, for example as WebP/JPEG under ${MAX_IMAGE_INPUT_DATA_URL_CHARS_LABEL} total base64 data-url chars.`);
+	if (totalDataUrlChars > MAX_IMAGE_INPUT_DATA_URL_CHARS) throw new Error(`qgrid: image inputs are too large for JSON transport (${totalDataUrlChars} total chars). Compress or resize them before passing them to imageGeneration, for example as WebP/JPEG under ${MAX_IMAGE_INPUT_DATA_URL_CHARS_LABEL} total base64 data-url chars.`);
 }
 function warnDroppedImages(droppedImageCount) {
 	if (droppedImageCount <= 0) return;
@@ -659,7 +659,39 @@ function qgrid(modelId, config) {
 	const serverUrl = config?.serverUrl ?? process.env.QGRID_URL ?? "http://localhost:44900";
 	const effort = config?.defaultEffort ?? "low";
 	const projectName = config?.projectName ?? process.env.QGRID_PROJECT_NAME;
-	let clientRun = null;
+	const clientRuns = /* @__PURE__ */ new Map();
+	let anonymousClientRunId = 0;
+	function clientRunKey(runContext) {
+		if (runContext.requestLogId !== void 0) return `request:${runContext.requestLogId}`;
+		if (runContext.threadCoord) {
+			const { workerId, epoch, threadId } = runContext.threadCoord;
+			return `thread:${workerId}:${epoch}:${threadId}`;
+		}
+		anonymousClientRunId++;
+		return `anonymous:${anonymousClientRunId}`;
+	}
+	function rememberClientRun(runContext, toolCallIds) {
+		if (toolCallIds.length === 0) return;
+		clientRuns.set(clientRunKey(runContext), {
+			runContext,
+			pendingToolCallIds: new Set(toolCallIds)
+		});
+	}
+	function matchClientRun(messages) {
+		const toolResults = extractToolResultsFromHistory(messages);
+		if (toolResults.length === 0) return null;
+		const resultIds = new Set(toolResults.map((r) => r.callId));
+		for (const [key, run] of clientRuns) if (run.pendingToolCallIds.size > 0 && [...run.pendingToolCallIds].every((id) => resultIds.has(id))) return {
+			key,
+			runContext: run.runContext,
+			toolResultsPayload: toolResults.filter((r) => run.pendingToolCallIds.has(r.callId)).map((r) => ({
+				toolCallId: r.callId,
+				output: r.result
+			}))
+		};
+		if (clientRuns.size > 0) console.warn(PENDING_TOOL_RESULTS_WARNING);
+		return null;
+	}
 	return {
 		specificationVersion: "v3",
 		provider: QGRID_PROVIDER_NAME,
@@ -687,20 +719,11 @@ function qgrid(modelId, config) {
 			let runContext;
 			let toolResultsPayload;
 			let logMode;
-			if (clientRun) {
-				const toolResults = extractToolResultsFromHistory(options.prompt);
-				const resultIds = new Set(toolResults.map((r) => r.callId));
-				if (clientRun.pendingToolCallIds.size > 0 && [...clientRun.pendingToolCallIds].every((id) => resultIds.has(id))) {
-					runContext = clientRun.runContext;
-					toolResultsPayload = toolResults.filter((r) => clientRun.pendingToolCallIds.has(r.callId)).map((r) => ({
-						toolCallId: r.callId,
-						output: r.result
-					}));
-					logMode = "run";
-				} else {
-					console.warn(PENDING_TOOL_RESULTS_WARNING);
-					clientRun = null;
-				}
+			const matchedClientRun = matchClientRun(options.prompt);
+			if (matchedClientRun) {
+				runContext = matchedClientRun.runContext;
+				toolResultsPayload = matchedClientRun.toolResultsPayload;
+				logMode = "run";
 			}
 			const storedCoord = reuseSessionKey ? getThreadCoord(reuseSessionKey) : void 0;
 			if (!runContext && storedCoord) runContext = { threadCoord: storedCoord };
@@ -765,11 +788,8 @@ function qgrid(modelId, config) {
 				text: data.text
 			});
 			if (imageGeneration && !content.some((c) => c.type === "file")) throw new Error("qgrid: imageGeneration was requested but the response contained no image — the server may be an older version that does not support image generation.");
-			if (data.runContext && finishReason.unified === "tool-calls") clientRun = {
-				runContext: data.runContext,
-				pendingToolCallIds: new Set(content.filter((c) => c.type === "tool-call").map((c) => c.toolCallId))
-			};
-			else clientRun = null;
+			if (matchedClientRun) clientRuns.delete(matchedClientRun.key);
+			if (data.runContext && finishReason.unified === "tool-calls") rememberClientRun(data.runContext, content.filter((c) => c.type === "tool-call").map((c) => c.toolCallId));
 			if (reuseSessionKey && data.runContext?.threadCoord) setThreadCoord(reuseSessionKey, data.runContext.threadCoord);
 			return {
 				content,
@@ -803,20 +823,11 @@ function qgrid(modelId, config) {
 			let runContext;
 			let toolResultsPayload;
 			let logMode;
-			if (clientRun) {
-				const toolResults = extractToolResultsFromHistory(options.prompt);
-				const resultIds = new Set(toolResults.map((r) => r.callId));
-				if (clientRun.pendingToolCallIds.size > 0 && [...clientRun.pendingToolCallIds].every((id) => resultIds.has(id))) {
-					runContext = clientRun.runContext;
-					toolResultsPayload = toolResults.filter((r) => clientRun.pendingToolCallIds.has(r.callId)).map((r) => ({
-						toolCallId: r.callId,
-						output: r.result
-					}));
-					logMode = "run";
-				} else {
-					console.warn(PENDING_TOOL_RESULTS_WARNING);
-					clientRun = null;
-				}
+			const matchedClientRun = matchClientRun(options.prompt);
+			if (matchedClientRun) {
+				runContext = matchedClientRun.runContext;
+				toolResultsPayload = matchedClientRun.toolResultsPayload;
+				logMode = "run";
 			}
 			const storedCoord = reuseSessionKey ? getThreadCoord(reuseSessionKey) : void 0;
 			if (!runContext && storedCoord) runContext = { threadCoord: storedCoord };
@@ -883,11 +894,8 @@ function qgrid(modelId, config) {
 							textStarted = false;
 						}
 						const done = event.data;
-						if (done.runContext && done.finishReason === "tool-calls") clientRun = {
-							runContext: done.runContext,
-							pendingToolCallIds: new Set((done.content ?? []).filter((c) => c.type === "tool-call").map((c) => c.toolCallId))
-						};
-						else clientRun = null;
+						if (matchedClientRun) clientRuns.delete(matchedClientRun.key);
+						if (done.runContext && done.finishReason === "tool-calls") rememberClientRun(done.runContext, (done.content ?? []).filter((c) => c.type === "tool-call").map((c) => c.toolCallId));
 						if (reuseSessionKey && done.runContext?.threadCoord) setThreadCoord(reuseSessionKey, done.runContext.threadCoord);
 						if (done.content) {
 							for (const item of done.content) if (item.type === "text" && !deltaTextEmitted) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@cartanova/qgrid-ai-sdk",
-  "version": "2.4.1",
+  "version": "2.4.3",
   "description": "AI SDK LanguageModelV3 provider for qgrid",
   "repository": {
     "type": "git",