feat(ai): add claude-sonnet-4-6 to Antigravity, fix Claude thinking header detection, bump UA to 1.18.4

closes #1859
2026-03-06 13:27:09 +01:00
parent b13e080113
commit e4172e68d0
5 changed files with 120 additions and 10 deletions
--- a/packages/ai/CHANGELOG.md
+++ b/packages/ai/CHANGELOG.md
@@ -2,6 +2,15 @@

 ## [Unreleased]

+### Added
+
+- Added `claude-sonnet-4-6` model for the `google-antigravity` provider ([#1859](https://github.com/badlogic/pi-mono/issues/1859)).
+- Bumped default Antigravity User-Agent version to `1.18.4` ([#1859](https://github.com/badlogic/pi-mono/issues/1859)).
+
+### Fixed
+
+- Fixed Antigravity Claude thinking beta header detection to use provider and model capability instead of `-thinking` suffix, so models like `claude-sonnet-4-6` receive the header correctly ([#1859](https://github.com/badlogic/pi-mono/issues/1859)).
+
 ## [0.56.2] - 2026-03-05

 ### Added
--- a/packages/ai/scripts/generate-models.ts
+++ b/packages/ai/scripts/generate-models.ts
@@ -1221,6 +1221,18 @@ async function generateModels() {
 			contextWindow: 200000,
 			maxTokens: 128000,
 		},
+		{
+			id: "claude-sonnet-4-6",
+			name: "Claude Sonnet 4.6 (Antigravity)",
+			api: "google-gemini-cli",
+			provider: "google-antigravity",
+			baseUrl: ANTIGRAVITY_ENDPOINT,
+			reasoning: true,
+			input: ["text", "image"],
+			cost: { input: 3, output: 15, cacheRead: 0.3, cacheWrite: 3.75 },
+			contextWindow: 200000,
+			maxTokens: 64000,
+		},
 		{
 			id: "gpt-oss-120b-medium",
 			name: "GPT-OSS 120B Medium (Antigravity)",
--- a/packages/ai/src/models.generated.ts
+++ b/packages/ai/src/models.generated.ts
@@ -2304,6 +2304,23 @@ export const MODELS = {
 			contextWindow: 272000,
 			maxTokens: 128000,
 		} satisfies Model<"azure-openai-responses">,
+		"gpt-5.4-pro": {
+			id: "gpt-5.4-pro",
+			name: "GPT-5.4 Pro",
+			api: "azure-openai-responses",
+			provider: "azure-openai-responses",
+			baseUrl: "",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 30,
+				output: 180,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 1050000,
+			maxTokens: 128000,
+		} satisfies Model<"azure-openai-responses">,
 		"o1": {
 			id: "o1",
 			name: "o1",
@@ -2880,7 +2897,7 @@ export const MODELS = {
 		} satisfies Model<"openai-responses">,
 		"gpt-5.3-codex": {
 			id: "gpt-5.3-codex",
-			name: "GPT-5.3 Codex",
+			name: "GPT-5.3-Codex",
 			api: "openai-responses",
 			provider: "github-copilot",
 			baseUrl: "https://api.individual.githubcopilot.com",
@@ -2893,7 +2910,25 @@ export const MODELS = {
 				cacheRead: 0,
 				cacheWrite: 0,
 			},
-			contextWindow: 272000,
+			contextWindow: 400000,
+			maxTokens: 128000,
+		} satisfies Model<"openai-responses">,
+		"gpt-5.4": {
+			id: "gpt-5.4",
+			name: "GPT-5.4",
+			api: "openai-responses",
+			provider: "github-copilot",
+			baseUrl: "https://api.individual.githubcopilot.com",
+			headers: {"User-Agent":"GitHubCopilotChat/0.35.0","Editor-Version":"vscode/1.107.0","Editor-Plugin-Version":"copilot-chat/0.35.0","Copilot-Integration-Id":"vscode-chat"},
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0,
+				output: 0,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 400000,
 			maxTokens: 128000,
 		} satisfies Model<"openai-responses">,
 		"grok-code-fast-1": {
@@ -3395,6 +3430,23 @@ export const MODELS = {
 			contextWindow: 200000,
 			maxTokens: 64000,
 		} satisfies Model<"google-gemini-cli">,
+		"claude-sonnet-4-6": {
+			id: "claude-sonnet-4-6",
+			name: "Claude Sonnet 4.6 (Antigravity)",
+			api: "google-gemini-cli",
+			provider: "google-antigravity",
+			baseUrl: "https://daily-cloudcode-pa.sandbox.googleapis.com",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 3,
+				output: 15,
+				cacheRead: 0.3,
+				cacheWrite: 3.75,
+			},
+			contextWindow: 200000,
+			maxTokens: 64000,
+		} satisfies Model<"google-gemini-cli">,
 		"gemini-3-flash": {
 			id: "gemini-3-flash",
 			name: "Gemini 3 Flash (Antigravity)",
@@ -5454,6 +5506,23 @@ export const MODELS = {
 			contextWindow: 272000,
 			maxTokens: 128000,
 		} satisfies Model<"openai-responses">,
+		"gpt-5.4-pro": {
+			id: "gpt-5.4-pro",
+			name: "GPT-5.4 Pro",
+			api: "openai-responses",
+			provider: "openai",
+			baseUrl: "https://api.openai.com/v1",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 30,
+				output: 180,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 1050000,
+			maxTokens: 128000,
+		} satisfies Model<"openai-responses">,
 		"o1": {
 			id: "o1",
 			name: "o1",
@@ -6172,6 +6241,23 @@ export const MODELS = {
 			contextWindow: 272000,
 			maxTokens: 128000,
 		} satisfies Model<"openai-responses">,
+		"gpt-5.4-pro": {
+			id: "gpt-5.4-pro",
+			name: "GPT-5.4 Pro",
+			api: "openai-responses",
+			provider: "opencode",
+			baseUrl: "https://opencode.ai/zen/v1",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 30,
+				output: 180,
+				cacheRead: 30,
+				cacheWrite: 0,
+			},
+			contextWindow: 1050000,
+			maxTokens: 128000,
+		} satisfies Model<"openai-responses">,
 		"kimi-k2.5": {
 			id: "kimi-k2.5",
 			name: "Kimi K2.5",
@@ -10350,9 +10436,9 @@ export const MODELS = {
 			reasoning: true,
 			input: ["text"],
 			cost: {
-				input: 0.3,
-				output: 1.4,
-				cacheRead: 0.15,
+				input: 0.38,
+				output: 1.9800000000000002,
+				cacheRead: 0.19,
 				cacheWrite: 0,
 			},
 			contextWindow: 202752,
--- a/packages/ai/src/providers/google-gemini-cli.ts
+++ b/packages/ai/src/providers/google-gemini-cli.ts
@@ -77,7 +77,7 @@ const GEMINI_CLI_HEADERS = {
 };

 // Headers for Antigravity (sandbox endpoint) - requires specific User-Agent
-const DEFAULT_ANTIGRAVITY_VERSION = "1.18.3";
+const DEFAULT_ANTIGRAVITY_VERSION = "1.18.4";

 function getAntigravityHeaders() {
 	const version = process.env.PI_AI_ANTIGRAVITY_VERSION || DEFAULT_ANTIGRAVITY_VERSION;
@@ -203,9 +203,8 @@ export function extractRetryDelay(errorText: string, response?: Response | Heade
 	return undefined;
 }

-function isClaudeThinkingModel(modelId: string): boolean {
-	const normalized = modelId.toLowerCase();
-	return normalized.includes("claude") && normalized.includes("thinking");
+function needsClaudeThinkingBetaHeader(model: Model<"google-gemini-cli">): boolean {
+	return model.provider === "google-antigravity" && model.id.startsWith("claude-") && model.reasoning;
 }

 function isGemini3ProModel(modelId: string): boolean {
@@ -379,7 +378,7 @@ export const streamGoogleGeminiCli: StreamFunction<"google-gemini-cli", GoogleGe
 				"Content-Type": "application/json",
 				Accept: "text/event-stream",
 				...headers,
-				...(isClaudeThinkingModel(model.id) ? { "anthropic-beta": CLAUDE_THINKING_BETA_HEADER } : {}),
+				...(needsClaudeThinkingBetaHeader(model) ? { "anthropic-beta": CLAUDE_THINKING_BETA_HEADER } : {}),
 				...options?.headers,
 			};
 			const requestBodyJson = JSON.stringify(requestBody);
--- a/packages/coding-agent/CHANGELOG.md
+++ b/packages/coding-agent/CHANGELOG.md
@@ -2,6 +2,10 @@

 ## [Unreleased]

+### Added
+
+- Added `claude-sonnet-4-6` model for the `google-antigravity` provider ([#1859](https://github.com/badlogic/pi-mono/issues/1859)).
+
 ### Fixed

 - Fixed `/new` leaving startup header content, including the changelog, visible after starting a fresh session ([#1880](https://github.com/badlogic/pi-mono/issues/1880))