feat: implement thinking for some more copilot models (#234)

Signed-off-by: StarLight842 <mail@aadishv.dev>

packages/ai/scripts/generate-models.ts

@@ -318,7 +318,7 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
 				if (m.status === "deprecated") continue;
 
 				// gpt-5 models require responses API, others use completions
-				const needsResponsesApi = modelId.startsWith("gpt-5");
+				const needsResponsesApi = modelId.startsWith("gpt-5") || modelId.startsWith("oswe");
 
 				const copilotModel: Model<any> = {
 					id: modelId,

packages/ai/src/models.generated.ts

@@ -2178,6 +2178,23 @@ export const MODELS = {
 			contextWindow: 128000,
 			maxTokens: 16384,
 		} satisfies Model<"openai-completions">,
+		"devstral-2512": {
+			id: "devstral-2512",
+			name: "Devstral 2",
+			api: "openai-completions",
+			provider: "mistral",
+			baseUrl: "https://api.mistral.ai/v1",
+			reasoning: false,
+			input: ["text"],
+			cost: {
+				input: 0,
+				output: 0,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 262144,
+			maxTokens: 262144,
+		} satisfies Model<"openai-completions">,
 		"ministral-3b-latest": {
 			id: "ministral-3b-latest",
 			name: "Ministral 3B",
@@ -2238,8 +2255,8 @@ export const MODELS = {
 			reasoning: false,
 			input: ["text", "image"],
 			cost: {
-				input: 0.1,
+				input: 0,
-				output: 0.3,
+				output: 0,
 				cacheRead: 0,
 				cacheWrite: 0,
 			},
@@ -2583,11 +2600,10 @@ export const MODELS = {
 		"oswe-vscode-prime": {
 			id: "oswe-vscode-prime",
 			name: "Raptor Mini (Preview)",
-			api: "openai-completions",
+			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"},
-			compat: {"supportsStore":false,"supportsDeveloperRole":false,"supportsReasoningEffort":false},
 			reasoning: true,
 			input: ["text", "image"],
 			cost: {
@@ -2598,7 +2614,7 @@ export const MODELS = {
 			},
 			contextWindow: 200000,
 			maxTokens: 64000,
-		} satisfies Model<"openai-completions">,
+		} satisfies Model<"openai-responses">,
 		"gpt-5.1-codex-mini": {
 			id: "gpt-5.1-codex-mini",
 			name: "GPT-5.1-Codex-mini",
@@ -2909,6 +2925,23 @@ export const MODELS = {
 			contextWindow: 256000,
 			maxTokens: 4096,
 		} satisfies Model<"openai-completions">,
+		"nvidia/nemotron-3-nano-30b-a3b": {
+			id: "nvidia/nemotron-3-nano-30b-a3b",
+			name: "NVIDIA: Nemotron 3 Nano 30B A3B",
+			api: "openai-completions",
+			provider: "openrouter",
+			baseUrl: "https://openrouter.ai/api/v1",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.06,
+				output: 0.24,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 262144,
+			maxTokens: 4096,
+		} satisfies Model<"openai-completions">,
 		"openai/gpt-5.2-chat": {
 			id: "openai/gpt-5.2-chat",
 			name: "OpenAI: GPT-5.2 Chat",
@@ -3190,13 +3223,13 @@ export const MODELS = {
 			reasoning: true,
 			input: ["text"],
 			cost: {
-				input: 0.26,
+				input: 0.24,
 				output: 0.38,
-				cacheRead: 0,
+				cacheRead: 0.19,
 				cacheWrite: 0,
 			},
 			contextWindow: 163840,
-			maxTokens: 65536,
+			maxTokens: 163840,
 		} satisfies Model<"openai-completions">,
 		"prime-intellect/intellect-3": {
 			id: "prime-intellect/intellect-3",
@@ -5451,13 +5484,13 @@ export const MODELS = {
 			reasoning: true,
 			input: ["text"],
 			cost: {
-				input: 0.15,
+				input: 0.19999999999999998,
-				output: 0.75,
+				output: 0.88,
-				cacheRead: 0,
+				cacheRead: 0.106,
 				cacheWrite: 0,
 			},
-			contextWindow: 8192,
+			contextWindow: 163840,
-			maxTokens: 7168,
+			maxTokens: 4096,
 		} satisfies Model<"openai-completions">,
 		"mistralai/mistral-small-3.1-24b-instruct:free": {
 			id: "mistralai/mistral-small-3.1-24b-instruct:free",

packages/ai/src/providers/openai-completions.ts

@@ -194,7 +194,7 @@ export const streamOpenAICompletions: StreamFunction<"openai-completions"> = (
 
 					// Some endpoints return reasoning in reasoning_content (llama.cpp),
 					// or reasoning (other openai compatible endpoints)
-					const reasoningFields = ["reasoning_content", "reasoning"];
+					const reasoningFields = ["reasoning_content", "reasoning", "reasoning_text"];
 					for (const field of reasoningFields) {
 						if (
 							(choice.delta as any)[field] !== null &&