@opencode/ai 0.0.0-dev-20125 → 0.0.0-dev-20127

package/dist/protocols/alibaba-chat.js

@@ -55,7 +55,10 @@ export const protocol = Protocol.make({
             return {
                 ...(yield* OpenAIChat.protocol.body.from(req)),
                 enable_thinking: opts.enableThinking,
-                thinking_budget: opts.thinkingBudget,
+                // Alibaba also rejects an explicit budget that is not below `max_completion_tokens`.
+                thinking_budget: opts.thinkingBudget === undefined
+                    ? undefined
+                    : ProviderShared.fitThinkingBudget(opts.thinkingBudget, req.generation?.maxTokens),
                 preserve_thinking: opts.preserveThinking,
                 clear_thinking: opts.clearThinking,
                 thinking: opts.thinking,

package/dist/protocols/alibaba-messages.js

@@ -22,15 +22,17 @@ export const protocol = Protocol.make({
         from: Effect.fn("AlibabaMessages.fromRequest")(function* (req) {
             const opts = yield* ProviderShared.validateWith(Schema.decodeUnknownEffect(Options))(req.providerOptions ?? {});
             // Model Studio accepts enabled thinking without Anthropic's mandatory token budget.
+            const body = yield* AnthropicMessages.protocol.body.from(LLMRequest.update(req, {
+                providerOptions: { ...req.providerOptions, thinking: undefined },
+            }));
+            const budget = opts.thinking?.budgetTokens ?? opts.thinking?.budget_tokens;
             return {
-                ...(yield* AnthropicMessages.protocol.body.from(LLMRequest.update(req, {
-                    providerOptions: { ...req.providerOptions, thinking: undefined },
-                }))),
+                ...body,
                 thinking: opts.thinking === undefined
                     ? undefined
                     : {
                         type: opts.thinking.type,
-                        budget_tokens: opts.thinking.budgetTokens ?? opts.thinking.budget_tokens,
+                        budget_tokens: budget === undefined ? undefined : ProviderShared.fitThinkingBudget(budget, body.max_tokens),
                     },
             };
         }),

package/dist/protocols/anthropic-messages.js

@@ -20,6 +20,7 @@ const ADAPTER = "anthropic-messages";
 export const DEFAULT_BASE_URL = "https://api.anthropic.com/v1";
 export const PATH = "/messages";
 export const DEFAULT_MAX_TOKENS = 32_000;
+const MIN_THINKING_BUDGET = 1_024;
 const DEFAULT_EFFORT = "high";
 const SSE_EVENTS = new Set([
     "message",
@@ -890,6 +891,13 @@ const applyThinkingBindingDefault = (model, thinking) => {
         },
     };
 };
+// Anthropic also requires an explicit thinking budget below `max_tokens` and at or above its minimum.
+const fitThinking = (thinking, maxTokens) => thinking?.type === "enabled"
+    ? {
+        ...thinking,
+        budget_tokens: ProviderShared.fitThinkingBudget(thinking.budget_tokens, maxTokens, MIN_THINKING_BUDGET),
+    }
+    : thinking;
 const fromRequest = Effect.fn("AnthropicMessages.fromRequest")(function* (request) {
     const options = yield* decodeOptions(request.providerOptions ?? {});
     const management = options.contextManagement;
@@ -920,6 +928,7 @@ const fromRequest = Effect.fn("AnthropicMessages.fromRequest")(function* (reques
         yield* Effect.logWarning(`Anthropic Messages: dropped ${breakpoints.dropped} cache breakpoint(s); the API allows at most ${ANTHROPIC_BREAKPOINT_CAP} per request.`);
     }
     const output_config = updates.effort === undefined && format === undefined ? undefined : { effort: updates.effort, format };
+    const maxTokens = generation?.maxTokens ?? DEFAULT_MAX_TOKENS;
     const body = {
         model: request.model.id,
         system,
@@ -927,12 +936,12 @@ const fromRequest = Effect.fn("AnthropicMessages.fromRequest")(function* (reques
         tools,
         tool_choice: toolChoice,
         stream: true,
-        max_tokens: generation?.maxTokens ?? DEFAULT_MAX_TOKENS,
+        max_tokens: maxTokens,
         temperature: generation?.temperature,
         top_p: generation?.topP,
         top_k: generation?.topK,
         stop_sequences: generation?.stop,
-        thinking: applyThinkingBindingDefault(request.model, options.thinking),
+        thinking: applyThinkingBindingDefault(request.model, fitThinking(options.thinking, maxTokens)),
         output_config,
         // top-level passthrough per SDK MessageCreateParamsBase:4638,4643,4649,4654,4670
         cache_control: options.cache_control ?? options.cacheControl,

package/dist/protocols/gemini.js

@@ -15,6 +15,8 @@ import { ToolSchemaProjection } from "./utils/tool-schema.js";
 const ADAPTER = "gemini";
 // Google documents this sentinel for replaying Gemini 3 function calls after their original signature was lost.
 const SKIP_THOUGHT_SIGNATURE_VALIDATOR = "skip_thought_signature_validator";
+// Gemini 2.5 rejects a budget under the model's minimum: 512 on Flash-Lite, the highest, and 128 on Pro.
+const MIN_THINKING_BUDGET = 512;
 export const DEFAULT_BASE_URL = "https://generativelanguage.googleapis.com/v1beta";
 // Gemini 3 rejects replayed function calls without a thought signature. Google's SDKs avoid that in normal chats by
 // retaining complete model responses, but OpenCode reconstructs durable history and may encounter an unsigned call
@@ -366,9 +368,16 @@ const fromRequest = Effect.fn("Gemini.fromRequest")(function* (request) {
         presencePenalty: generation?.presencePenalty,
         seed: generation?.seed,
         stopSequences: generation?.stop,
+        // Gemini accepts a budget above `maxOutputTokens`, but thinking then leaves the answer empty.
         thinkingConfig: options.thinkingConfig === undefined
             ? undefined
-            : { ...options.thinkingConfig, includeThoughts: options.thinkingConfig.includeThoughts ?? true },
+            : {
+                ...options.thinkingConfig,
+                includeThoughts: options.thinkingConfig.includeThoughts ?? true,
+                thinkingBudget: options.thinkingConfig.thinkingBudget === undefined
+                    ? undefined
+                    : ProviderShared.fitThinkingBudget(options.thinkingConfig.thinkingBudget, generation?.maxTokens, MIN_THINKING_BUDGET),
+            },
     };
     return {
         cachedContent: options.cachedContent,

package/dist/protocols/shared.d.ts

@@ -58,6 +58,12 @@ export declare const subtractTokens: (total: number | undefined, subtrahend: num
  * (e.g. Anthropic, whose `input_tokens` is already non-cached only).
  */
 export declare const sumTokens: (...values: ReadonlyArray<number | undefined>) => number | undefined;
+/**
+ * Caps an explicit thinking budget at half the output limit. Thinking counts against the output limit, so a budget
+ * near it leaves the answer, a tool call, or a summary without room. Smaller budgets, special values such as `-1` and
+ * `0`, and requests without an output limit pass through unchanged.
+ */
+export declare const fitThinkingBudget: (budget: number, maxTokens: number | undefined, minimum?: number) => number;
 export declare const eventError: (route: string, message: string, body?: string, cause?: unknown) => AIError;
 export declare const parseJson: (route: string, input: string, message: string) => Effect.Effect<unknown, AIError, never>;
 /**

package/dist/protocols/shared.js

@@ -75,6 +75,12 @@ export const sumTokens = (...values) => {
         return undefined;
     return values.reduce((acc, value) => acc + (value ?? 0), 0);
 };
+/**
+ * Caps an explicit thinking budget at half the output limit. Thinking counts against the output limit, so a budget
+ * near it leaves the answer, a tool call, or a summary without room. Smaller budgets, special values such as `-1` and
+ * `0`, and requests without an output limit pass through unchanged.
+ */
+export const fitThinkingBudget = (budget, maxTokens, minimum = 1) => maxTokens === undefined || budget <= maxTokens / 2 ? budget : Math.max(minimum, Math.floor(maxTokens / 2));
 export const eventError = (route, message, body, cause) => new AIError({
     reason: new InvalidProviderOutputError({ route, message, body, cause }),
 });

package/package.json

@@ -1,6 +1,6 @@
 {
   "$schema": "https://json.schemastore.org/package.json",
-  "version": "0.0.0-dev-20125",
+  "version": "0.0.0-dev-20127",
   "name": "@opencode/ai",
   "type": "module",
   "license": "MIT",
@@ -34,7 +34,7 @@
   "devDependencies": {
     "@clack/prompts": "1.0.0-alpha.1",
     "@effect/platform-node": "4.0.0-rc.112",
-    "@opencode/http-recorder": "0.0.0-dev-20125",
+    "@opencode/http-recorder": "0.0.0-dev-20127",
     "@tsconfig/bun": "1.0.9",
     "@types/bun": "1.4.0",
     "@typescript/native-preview": "7.0.0-dev.20251207.1",
@@ -44,7 +44,7 @@
     "@aws-sdk/credential-providers": "3.1057.0",
     "@smithy/eventstream-codec": "4.2.14",
     "@smithy/util-utf8": "4.2.2",
-    "@opencode/schema": "0.0.0-dev-20125",
+    "@opencode/schema": "0.0.0-dev-20127",
     "aws4fetch": "1.0.20",
     "effect": "4.0.0-rc.112",
     "google-auth-library": "10.5.0"