@opencode/ai 0.0.0-dev-20127 → 0.0.0-dev-20131

package/README.md

@@ -72,10 +72,11 @@ helpers; `ai.file` and `ai.write` load `node:fs/promises` on first use, so no Ef
 import { AI } from "@opencode/ai/promise"
 
 const ai = AI.make()
-const text = await ai.llm.generate({ model: openai.responses("gpt-4o-mini"), prompt: "Say hello." })
+const request = ai.llm.request({ model: openai.responses("gpt-4o-mini"), prompt: "Say hello." })
+const text = await ai.llm.generate(request)
 const generated = await ai.image.generate({ model: openai.image("gpt-image-2"), prompt: "A lighthouse" })
 await ai.write(generated.image, "./lighthouse.png") // also ai.file(path), ai.bytes(asset), ai.base64(asset), ai.materialize(asset)
-for await (const event of ai.llm.stream({ model: openai.responses("gpt-4o-mini"), prompt: "Stream hello." })) {
+for await (const event of ai.llm.stream(request)) {
   // LLMEvent
 }
 await ai.dispose()

package/dist/llm.d.ts

@@ -14,6 +14,7 @@ export type RequestInput<SelectedLanguageModel extends LanguageModel = LanguageM
     readonly providerOptions?: NoInfer<LanguageModelProviderOptions<SelectedLanguageModel>>;
     readonly http?: HttpOptions.Input;
 };
+/** Run an `LLMRequest` built with `LLM.request(...)`. */
 export declare const generate: typeof import("./route/client.js").generate;
 export declare const stream: typeof import("./route/client.js").stream;
 export declare const request: <const SelectedLanguageModel extends LanguageModel>(input: RequestInput<SelectedLanguageModel>) => LLMRequest<SelectedLanguageModel>;

package/dist/llm.js

@@ -2,6 +2,7 @@ import { Effect, JsonSchema, Schema } from "effect";
 import { LLMClient, Service } from "./route/client.js";
 import { GenerationOptions, HttpOptions, InvalidProviderOutputError, AIError, LLMEvent, LLMRequest, LLMResponse, Message, LanguageModel, SystemPart, ToolChoice, ToolEntry, } from "./schema/index.js";
 import { make as makeTool, toDefinitions } from "./tool.js";
+/** Run an `LLMRequest` built with `LLM.request(...)`. */
 export const generate = LLMClient.generate;
 export const stream = LLMClient.stream;
 export const request = (input) => {

package/dist/promise.d.ts

@@ -2,10 +2,10 @@ import { Effect, Layer } from "effect";
 import { AIClient } from "./ai-client.js";
 import type { AwaitOptions, Event, Snapshot } from "./generation.js";
 import { ImageModel, ImageRequest, type ImageOptions, type ImageRequestInput } from "./image.js";
+import { LLM } from "./index.js";
 import { Media } from "./media.js";
 import { RequestExecutor } from "./route/executor.js";
-import { LanguageModel, LLMRequest } from "./schema/index.js";
-import type { RequestInput } from "./llm.js";
+import { type LLMRequest } from "./schema/index.js";
 import { SpeechModel, SpeechRequest, type SpeechRequestInput } from "./speech.js";
 import { TranscriptionModel, TranscriptionRequest, type TranscriptionOptions, type TranscriptionRequestInput } from "./transcription.js";
 import { VideoModel, VideoRequest, type VideoOptions, type VideoRequestInput } from "./video.js";
@@ -49,9 +49,9 @@ export declare const make: (options?: Options) => {
     /** Write an asset's bytes like `Media.write`, downloading `url` sources through the executor. */
     write: (asset: Media.Asset, path: string, options?: RunOptions) => Promise<void>;
     llm: {
-        request: <const SelectedLanguageModel extends LanguageModel>(input: RequestInput<SelectedLanguageModel>) => LLMRequest<SelectedLanguageModel>;
-        generate: <const Model extends LanguageModel>(input: RequestInput<Model> | LLMRequest, options?: RunOptions) => Promise<import("./index.js").LLMResponse>;
-        stream: <const Model extends LanguageModel>(input: RequestInput<Model> | LLMRequest, options?: RunOptions) => AsyncIterable<({
+        request: <const SelectedLanguageModel extends import("./index.js").LanguageModel>(input: LLM.RequestInput<SelectedLanguageModel>) => LLMRequest<SelectedLanguageModel>;
+        generate: (request: LLMRequest, options?: RunOptions) => Promise<import("./index.js").LLMResponse>;
+        stream: (request: LLMRequest, options?: RunOptions) => AsyncIterable<({
             readonly type: "compaction";
             readonly provider: string & import("effect/Brand").Brand<"AI.ProviderID">;
             readonly id?: string | undefined;
@@ -528,9 +528,9 @@ export declare const ai: {
     /** Write an asset's bytes like `Media.write`, downloading `url` sources through the executor. */
     write: (asset: Media.Asset, path: string, options?: RunOptions) => Promise<void>;
     llm: {
-        request: <const SelectedLanguageModel extends LanguageModel>(input: RequestInput<SelectedLanguageModel>) => LLMRequest<SelectedLanguageModel>;
-        generate: <const Model extends LanguageModel>(input: RequestInput<Model> | LLMRequest, options?: RunOptions) => Promise<import("./index.js").LLMResponse>;
-        stream: <const Model extends LanguageModel>(input: RequestInput<Model> | LLMRequest, options?: RunOptions) => AsyncIterable<({
+        request: <const SelectedLanguageModel extends import("./index.js").LanguageModel>(input: LLM.RequestInput<SelectedLanguageModel>) => LLMRequest<SelectedLanguageModel>;
+        generate: (request: LLMRequest, options?: RunOptions) => Promise<import("./index.js").LLMResponse>;
+        stream: (request: LLMRequest, options?: RunOptions) => AsyncIterable<({
             readonly type: "compaction";
             readonly provider: string & import("effect/Brand").Brand<"AI.ProviderID">;
             readonly id?: string | undefined;

package/dist/promise.js

@@ -3,9 +3,8 @@ import { AIClient } from "./ai-client.js";
 import { Image, ImageModel, ImageRequest } from "./image.js";
 import { LLM } from "./index.js";
 import { Media } from "./media.js";
-import { tryRequest } from "./media-model.js";
 import { RequestExecutor } from "./route/executor.js";
-import { AIError, InvalidRequestError, LanguageModel, LLMRequest } from "./schema/index.js";
+import { AIError, InvalidRequestError } from "./schema/index.js";
 import { Speech, SpeechModel, SpeechRequest } from "./speech.js";
 import { Transcription, TranscriptionModel, TranscriptionRequest, } from "./transcription.js";
 import { fileMediaType } from "./utils/media-type.js";
@@ -36,7 +35,6 @@ export const make = (options = {}) => {
         cancel: (options) => run(generation.cancel(), options),
     });
     // The typed `generate`/`stream` overloads take a concrete input or a request, not the union; normalize once here.
-    const llmRequest = (input) => input instanceof LLMRequest ? Effect.succeed(input) : tryRequest(() => LLM.request(input));
     const imageRequest = (input) => input instanceof ImageRequest ? input : Image.request(input);
     const videoRequest = (input) => input instanceof VideoRequest ? input : Video.request(input);
     const speechRequest = (input) => input instanceof SpeechRequest ? input : Speech.request(input);
@@ -69,8 +67,8 @@ export const make = (options = {}) => {
         },
         llm: {
             request: LLM.request,
-            generate: (input, options) => run(Effect.flatMap(llmRequest(input), LLM.generate), options),
-            stream: (input, options) => iterate(Stream.unwrap(Effect.map(llmRequest(input), LLM.stream)), options),
+            generate: (request, options) => run(LLM.generate(request), options),
+            stream: (request, options) => iterate(LLM.stream(request), options),
         },
         image: {
             request: Image.request,

package/dist/providers/openrouter.js

@@ -7,7 +7,7 @@ import { HttpOptions, ProviderID } from "../schema/index.js";
 import { SystemOne } from "../experimental/system-one.js";
 import { OpenAIChat } from "../protocols/openai-chat.js";
 import { newBreakpoints, ttlBucket } from "../protocols/utils/cache.js";
-import { isRecord } from "../protocols/shared.js";
+import { isRecord, ProviderShared } from "../protocols/shared.js";
 export const id = ProviderID.make("openrouter");
 const baseURL = "https://openrouter.ai/api/v1";
 const ADAPTER = "openrouter";
@@ -41,7 +41,7 @@ export const protocol = Protocol.make({
             return {
                 ...body,
                 messages,
-                ...bodyOptions(request.providerOptions),
+                ...bodyOptions(request.providerOptions, request.generation?.maxTokens),
             };
         })),
     },
@@ -59,7 +59,12 @@ const cacheControl = () => {
         };
     };
 };
-const bodyOptions = (input) => {
+// OpenRouter forwards `reasoning.max_tokens` as the upstream thinking budget. Upstreams such as Anthropic and Alibaba
+// reject one that is not below the output limit; 1,024 is Anthropic's minimum budget.
+const fitReasoning = (reasoning, maxTokens) => typeof reasoning.max_tokens === "number"
+    ? { ...reasoning, max_tokens: ProviderShared.fitThinkingBudget(reasoning.max_tokens, maxTokens, 1_024) }
+    : reasoning;
+const bodyOptions = (input, maxTokens) => {
     const openrouter = isRecord(input) ? input : {};
     const { usage, models, provider, plugins, web_search_options, debug, user, reasoning, promptCacheKey, ...options } = openrouter;
     return {
@@ -77,7 +82,7 @@ const bodyOptions = (input) => {
         ...(isRecord(web_search_options) ? { web_search_options } : {}),
         ...(isRecord(debug) ? { debug } : {}),
         ...(typeof user === "string" ? { user } : {}),
-        ...(isRecord(reasoning) ? { reasoning } : {}),
+        ...(isRecord(reasoning) ? { reasoning: fitReasoning(reasoning, maxTokens) } : {}),
     };
 };
 export const route = Route.make({

package/package.json

@@ -1,6 +1,6 @@
 {
   "$schema": "https://json.schemastore.org/package.json",
-  "version": "0.0.0-dev-20127",
+  "version": "0.0.0-dev-20131",
   "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-20127",
+    "@opencode/http-recorder": "0.0.0-dev-20131",
     "@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-20127",
+    "@opencode/schema": "0.0.0-dev-20131",
     "aws4fetch": "1.0.20",
     "effect": "4.0.0-rc.112",
     "google-auth-library": "10.5.0"