feat(ai): add Cloudflare Workers AI as a provider (#3851)

* feat(ai): add Cloudflare Workers AI as a provider Cloudflare Workers AI hosts open-weight LLMs (Kimi K2.6, GPT-OSS, GLM-4.7, Llama 4, Gemma 4, Nemotron 3) on Cloudflare's GPU network with an OpenAI-compatible endpoint. Reuses the openai-completions API protocol; the per-account URL contains a {CLOUDFLARE_ACCOUNT_ID} placeholder resolved at request time by a small helper. Pi automatically sets x-session-affinity for prefix caching: https://developers.cloudflare.com/workers-ai/features/prompt-caching/ Auth: CLOUDFLARE_API_KEY (matches pi's *_API_KEY convention) + CLOUDFLARE_ACCOUNT_ID. The User-Agent identifies traffic as 'pi-coding-agent' in Cloudflare analytics. Verified end-to-end against a real Cloudflare account: 17 e2e tests pass across stream/empty/tokens/unicode/tool-call-without-result/ total-tokens against @cf/moonshotai/kimi-k2.6. Cloudflare AI Gateway is a separate, larger change (it requires routing through provider-specific subpaths with the matching API protocol per upstream) and will land in a follow-up PR. * refactor(ai): move Cloudflare User-Agent and session-affinity flag to per-model metadata Instead of conditionally setting them in openai-completions.ts based on provider detection, declare them as model-level fields in the catalog (headers + compat). This is consistent with how the github-copilot and kimi-coding entries already declare their static headers. packages/ai/scripts/generate-models.ts: emit headers and compat fields on each cloudflare-workers-ai entry (CLOUDFLARE_STATIC_HEADERS). packages/ai/src/providers/openai-completions.ts: drop the isCloudflareProvider conditional that injected User-Agent and the isCloudflareWorkersAI override of sendSessionAffinityHeaders. packages/ai/src/models.generated.ts: re-spliced 8 cloudflare-workers-ai entries with headers + compat. Behavior is unchanged - verified via fetch interceptor that User-Agent and x-session-affinity / session_id / x-client-request-id are still sent on outbound requests. 5/5 e2e tests pass.
2026-04-27 17:41:54 -04:00
parent 31b48dec23
commit d6e08b3da0
19 changed files with 367 additions and 4 deletions
--- a/packages/ai/README.md
+++ b/packages/ai/README.md
@@ -57,6 +57,7 @@ Unified LLM API with automatic model discovery, provider configuration, token an
 - **Mistral**
 - **Groq**
 - **Cerebras**
+- **Cloudflare Workers AI**
 - **xAI**
 - **OpenRouter**
 - **Vercel AI Gateway**
@@ -844,7 +845,7 @@ const ollamaReasoningModel: Model<'openai-completions'> = {

 ### OpenAI Compatibility Settings

-The `openai-completions` API is implemented by many providers with minor differences. By default, the library auto-detects compatibility settings based on `baseUrl` for a small set of known OpenAI-compatible providers (Cerebras, xAI, Chutes, DeepSeek, zAi, OpenCode, etc.). For custom proxies or unknown endpoints, you can override these settings via the `compat` field. For `openai-responses` models, the compat field only supports Responses-specific flags.
+The `openai-completions` API is implemented by many providers with minor differences. By default, the library auto-detects compatibility settings based on `baseUrl` for a small set of known OpenAI-compatible providers (Cerebras, xAI, Chutes, DeepSeek, zAi, OpenCode, Cloudflare Workers AI, etc.). For custom proxies or unknown endpoints, you can override these settings via the `compat` field. For `openai-responses` models, the compat field only supports Responses-specific flags.

 ```typescript
 interface OpenAICompletionsCompat {
@@ -1028,6 +1029,7 @@ In Node.js environments, you can set environment variables to avoid passing API
 | Mistral | `MISTRAL_API_KEY` |
 | Groq | `GROQ_API_KEY` |
 | Cerebras | `CEREBRAS_API_KEY` |
+| Cloudflare Workers AI | `CLOUDFLARE_API_KEY` + `CLOUDFLARE_ACCOUNT_ID` |
 | xAI | `XAI_API_KEY` |
 | Fireworks | `FIREWORKS_API_KEY` |
 | OpenRouter | `OPENROUTER_API_KEY` |
--- a/packages/ai/scripts/generate-models.ts
+++ b/packages/ai/scripts/generate-models.ts
@@ -3,6 +3,7 @@
 import { writeFileSync } from "fs";
 import { join, dirname } from "path";
 import { fileURLToPath } from "url";
+import { CLOUDFLARE_WORKERS_AI_BASE_URL } from "../src/providers/cloudflare.js";
 import {
 	Api,
 	type AnthropicMessagesCompat,
@@ -63,6 +64,10 @@ const KIMI_STATIC_HEADERS = {
 	"User-Agent": "KimiCLI/1.5",
 } as const;

+const CLOUDFLARE_STATIC_HEADERS = {
+	"User-Agent": "pi-coding-agent",
+} as const;
+
 const AI_GATEWAY_MODELS_URL = "https://ai-gateway.vercel.sh/v1";
 const AI_GATEWAY_BASE_URL = "https://ai-gateway.vercel.sh";
 const ZAI_TOOL_STREAM_UNSUPPORTED_MODELS = new Set(["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v"]);
@@ -376,6 +381,34 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
 			}
 		}

+		// Process Cloudflare Workers AI models
+		if (data["cloudflare-workers-ai"]?.models) {
+			for (const [modelId, model] of Object.entries(data["cloudflare-workers-ai"].models)) {
+				const m = model as ModelsDevModel;
+				if (m.tool_call !== true) continue;
+
+				models.push({
+					id: modelId,
+					name: m.name || modelId,
+					api: "openai-completions",
+					provider: "cloudflare-workers-ai",
+					baseUrl: CLOUDFLARE_WORKERS_AI_BASE_URL,
+					reasoning: m.reasoning === true,
+					input: m.modalities?.input?.includes("image") ? ["text", "image"] : ["text"],
+					cost: {
+						input: m.cost?.input || 0,
+						output: m.cost?.output || 0,
+						cacheRead: m.cost?.cache_read || 0,
+						cacheWrite: m.cost?.cache_write || 0,
+					},
+					contextWindow: m.limit?.context || 4096,
+					maxTokens: m.limit?.output || 4096,
+					headers: { ...CLOUDFLARE_STATIC_HEADERS },
+					compat: { sendSessionAffinityHeaders: true },
+				});
+			}
+		}
+
 		// Process xAi models
 		if (data.xai?.models) {
 			for (const [modelId, model] of Object.entries(data.xai.models)) {
--- a/packages/ai/src/env-api-keys.ts
+++ b/packages/ai/src/env-api-keys.ts
@@ -118,6 +118,7 @@ function getApiKeyEnvVars(provider: string): readonly string[] | undefined {
 		opencode: "OPENCODE_API_KEY",
 		"opencode-go": "OPENCODE_API_KEY",
 		"kimi-coding": "KIMI_API_KEY",
+		"cloudflare-workers-ai": "CLOUDFLARE_API_KEY",
 	};

 	const envVar = envMap[provider];
--- a/packages/ai/src/models.generated.ts
+++ b/packages/ai/src/models.generated.ts
@@ -15374,4 +15374,158 @@ export const MODELS = {
 			maxTokens: 131072,
 		} satisfies Model<"openai-completions">,
 	},
+	"cloudflare-workers-ai": {
+		"@cf/google/gemma-4-26b-a4b-it": {
+			id: "@cf/google/gemma-4-26b-a4b-it",
+			name: "Gemma 4 26B A4B IT",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0.1,
+				output: 0.3,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 16384,
+		} satisfies Model<"openai-completions">,
+		"@cf/meta/llama-4-scout-17b-16e-instruct": {
+			id: "@cf/meta/llama-4-scout-17b-16e-instruct",
+			name: "Llama 4 Scout 17B 16E Instruct",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: false,
+			input: ["text", "image"],
+			cost: {
+				input: 0.27,
+				output: 0.85,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 128000,
+			maxTokens: 16384,
+		} satisfies Model<"openai-completions">,
+		"@cf/moonshotai/kimi-k2.5": {
+			id: "@cf/moonshotai/kimi-k2.5",
+			name: "Kimi K2.5",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0.6,
+				output: 3,
+				cacheRead: 0.1,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 256000,
+		} satisfies Model<"openai-completions">,
+		"@cf/moonshotai/kimi-k2.6": {
+			id: "@cf/moonshotai/kimi-k2.6",
+			name: "Kimi K2.6",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0.95,
+				output: 4,
+				cacheRead: 0.16,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 256000,
+		} satisfies Model<"openai-completions">,
+		"@cf/nvidia/nemotron-3-120b-a12b": {
+			id: "@cf/nvidia/nemotron-3-120b-a12b",
+			name: "Nemotron 3 Super 120B",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.5,
+				output: 1.5,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 256000,
+		} satisfies Model<"openai-completions">,
+		"@cf/openai/gpt-oss-120b": {
+			id: "@cf/openai/gpt-oss-120b",
+			name: "GPT OSS 120B",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.35,
+				output: 0.75,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 128000,
+			maxTokens: 16384,
+		} satisfies Model<"openai-completions">,
+		"@cf/openai/gpt-oss-20b": {
+			id: "@cf/openai/gpt-oss-20b",
+			name: "GPT OSS 20B",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.2,
+				output: 0.3,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 128000,
+			maxTokens: 16384,
+		} satisfies Model<"openai-completions">,
+		"@cf/zai-org/glm-4.7-flash": {
+			id: "@cf/zai-org/glm-4.7-flash",
+			name: "GLM-4.7-Flash",
+			api: "openai-completions",
+			provider: "cloudflare-workers-ai",
+			baseUrl: "https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1",
+			headers: {"User-Agent":"pi-coding-agent"},
+			compat: {"sendSessionAffinityHeaders":true},
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.06,
+				output: 0.4,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 131072,
+			maxTokens: 131072,
+		} satisfies Model<"openai-completions">,
+	},
 } as const;
--- a/packages/ai/src/providers/cloudflare.ts
+++ b/packages/ai/src/providers/cloudflare.ts
@@ -0,0 +1,22 @@
+import type { Model } from "../types.js";
+
+/** Workers AI endpoint. `{CLOUDFLARE_ACCOUNT_ID}` is substituted at request time. */
+export const CLOUDFLARE_WORKERS_AI_BASE_URL =
+	"https://api.cloudflare.com/client/v4/accounts/{CLOUDFLARE_ACCOUNT_ID}/ai/v1";
+
+export function isCloudflareProvider(provider: string): boolean {
+	return provider === "cloudflare-workers-ai";
+}
+
+/** Substitute `{VAR}` placeholders in a Cloudflare baseUrl from process.env. */
+export function resolveCloudflareBaseUrl(model: Model<"openai-completions">): string {
+	const url = model.baseUrl;
+	if (!url.includes("{")) return url;
+	return url.replace(/\{([A-Z_][A-Z0-9_]*)\}/g, (_match, name: string) => {
+		const value = process.env[name];
+		if (!value) {
+			throw new Error(`${name} is required for provider ${model.provider} but is not set.`);
+		}
+		return value;
+	});
+}
--- a/packages/ai/src/providers/openai-completions.ts
+++ b/packages/ai/src/providers/openai-completions.ts
@@ -34,6 +34,7 @@ import { AssistantMessageEventStream } from "../utils/event-stream.js";
 import { headersToRecord } from "../utils/headers.js";
 import { parseStreamingJson } from "../utils/json-parse.js";
 import { sanitizeSurrogates } from "../utils/sanitize-unicode.js";
+import { isCloudflareProvider, resolveCloudflareBaseUrl } from "./cloudflare.js";
 import { buildCopilotDynamicHeaders, hasCopilotVisionInput } from "./github-copilot-headers.js";
 import { buildBaseOptions, clampReasoning } from "./simple-options.js";
 import { transformMessages } from "./transform-messages.js";
@@ -456,7 +457,7 @@ function createClient(

 	return new OpenAI({
 		apiKey,
-		baseURL: model.baseUrl,
+		baseURL: isCloudflareProvider(model.provider) ? resolveCloudflareBaseUrl(model) : model.baseUrl,
 		dangerouslyAllowBrowser: true,
 		defaultHeaders: headers,
 	});
@@ -1022,6 +1023,7 @@ function detectCompat(model: Model<"openai-completions">): ResolvedOpenAIComplet
 	const baseUrl = model.baseUrl;

 	const isZai = provider === "zai" || baseUrl.includes("api.z.ai");
+	const isCloudflareWorkersAI = provider === "cloudflare-workers-ai" || baseUrl.includes("api.cloudflare.com");

 	const isNonStandard =
 		provider === "cerebras" ||
@@ -1032,7 +1034,8 @@ function detectCompat(model: Model<"openai-completions">): ResolvedOpenAIComplet
 		baseUrl.includes("deepseek.com") ||
 		isZai ||
 		provider === "opencode" ||
-		baseUrl.includes("opencode.ai");
+		baseUrl.includes("opencode.ai") ||
+		isCloudflareWorkersAI;

 	const useMaxTokens = baseUrl.includes("chutes.ai");

--- a/packages/ai/src/types.ts
+++ b/packages/ai/src/types.ts
@@ -41,7 +41,8 @@ export type KnownProvider =
 	| "fireworks"
 	| "opencode"
 	| "opencode-go"
-	| "kimi-coding";
+	| "kimi-coding"
+	| "cloudflare-workers-ai";
 export type Provider = KnownProvider | string;

 export type ThinkingLevel = "minimal" | "low" | "medium" | "high" | "xhigh";
--- a/packages/ai/test/cross-provider-handoff.test.ts
+++ b/packages/ai/test/cross-provider-handoff.test.ts
@@ -84,6 +84,8 @@ const PROVIDER_MODEL_PAIRS: ProviderModelPair[] = [
 	{ provider: "xai", model: "grok-code-fast-1", label: "xai-grok-code-fast-1" },
 	// Cerebras
 	{ provider: "cerebras", model: "zai-glm-4.7", label: "cerebras-zai-glm-4.7" },
+	// Cloudflare Workers AI
+	{ provider: "cloudflare-workers-ai", model: "@cf/moonshotai/kimi-k2.6", label: "cloudflare-kimi-k2.6" },
 	// Groq
 	{ provider: "groq", model: "openai/gpt-oss-120b", label: "groq-gpt-oss-120b" },
 	// Hugging Face
--- a/packages/ai/test/empty.test.ts
+++ b/packages/ai/test/empty.test.ts
@@ -308,6 +308,29 @@ describe("AI Providers Empty Message Tests", () => {
 		});
 	});

+	describe.skipIf(!process.env.CLOUDFLARE_API_KEY || !process.env.CLOUDFLARE_ACCOUNT_ID)(
+		"Cloudflare Workers AI Provider Empty Messages",
+		() => {
+			const llm = getModel("cloudflare-workers-ai", "@cf/moonshotai/kimi-k2.6");
+
+			it("should handle empty content array", { retry: 3, timeout: 30000 }, async () => {
+				await testEmptyMessage(llm);
+			});
+
+			it("should handle empty string content", { retry: 3, timeout: 30000 }, async () => {
+				await testEmptyStringMessage(llm);
+			});
+
+			it("should handle whitespace-only content", { retry: 3, timeout: 30000 }, async () => {
+				await testWhitespaceOnlyMessage(llm);
+			});
+
+			it("should handle empty assistant message in conversation", { retry: 3, timeout: 30000 }, async () => {
+				await testEmptyAssistantMessage(llm);
+			});
+		},
+	);
+
 	describe.skipIf(!process.env.HF_TOKEN)("Hugging Face Provider Empty Messages", () => {
 		const llm = getModel("huggingface", "moonshotai/Kimi-K2.5");

--- a/packages/ai/test/stream.test.ts
+++ b/packages/ai/test/stream.test.ts
@@ -616,6 +616,33 @@ describe("Generate E2E Tests", () => {
 		});
 	});

+	describe.skipIf(!process.env.CLOUDFLARE_API_KEY || !process.env.CLOUDFLARE_ACCOUNT_ID)(
+		"Cloudflare Workers AI Provider (Kimi K2.6 via OpenAI Completions)",
+		() => {
+			const llm = getModel("cloudflare-workers-ai", "@cf/moonshotai/kimi-k2.6");
+
+			it("should complete basic text generation", { retry: 3 }, async () => {
+				await basicTextGeneration(llm);
+			});
+
+			it("should handle tool calling", { retry: 3 }, async () => {
+				await handleToolCall(llm);
+			});
+
+			it("should handle streaming", { retry: 3 }, async () => {
+				await handleStreaming(llm);
+			});
+
+			it("should handle thinking mode", { retry: 3 }, async () => {
+				await handleThinking(llm, { reasoningEffort: "medium" });
+			});
+
+			it("should handle multi-turn with thinking and tools", { retry: 3 }, async () => {
+				await multiTurn(llm, { reasoningEffort: "medium" });
+			});
+		},
+	);
+
 	describe.skipIf(!process.env.HF_TOKEN)("Hugging Face Provider (Kimi-K2.5 via OpenAI Completions)", () => {
 		const llm = getModel("huggingface", "moonshotai/Kimi-K2.5");

--- a/packages/ai/test/tokens.test.ts
+++ b/packages/ai/test/tokens.test.ts
@@ -159,6 +159,17 @@ describe("Token Statistics on Abort", () => {
 		});
 	});

+	describe.skipIf(!process.env.CLOUDFLARE_API_KEY || !process.env.CLOUDFLARE_ACCOUNT_ID)(
+		"Cloudflare Workers AI Provider",
+		() => {
+			const llm = getModel("cloudflare-workers-ai", "@cf/moonshotai/kimi-k2.6");
+
+			it("should include token stats when aborted mid-stream", { retry: 3, timeout: 30000 }, async () => {
+				await testTokensOnAbort(llm);
+			});
+		},
+	);
+
 	describe.skipIf(!process.env.HF_TOKEN)("Hugging Face Provider", () => {
 		const llm = getModel("huggingface", "moonshotai/Kimi-K2.5");

--- a/packages/ai/test/tool-call-without-result.test.ts
+++ b/packages/ai/test/tool-call-without-result.test.ts
@@ -168,6 +168,21 @@ describe("Tool Call Without Result Tests", () => {
 		});
 	});

+	describe.skipIf(!process.env.CLOUDFLARE_API_KEY || !process.env.CLOUDFLARE_ACCOUNT_ID)(
+		"Cloudflare Workers AI Provider",
+		() => {
+			const model = getModel("cloudflare-workers-ai", "@cf/moonshotai/kimi-k2.6");
+
+			it(
+				"should filter out tool calls without corresponding tool results",
+				{ retry: 3, timeout: 30000 },
+				async () => {
+					await testToolCallWithoutResult(model);
+				},
+			);
+		},
+	);
+
 	describe.skipIf(!process.env.HF_TOKEN)("Hugging Face Provider", () => {
 		const model = getModel("huggingface", "moonshotai/Kimi-K2.5");

--- a/packages/ai/test/total-tokens.test.ts
+++ b/packages/ai/test/total-tokens.test.ts
@@ -306,6 +306,34 @@ describe("totalTokens field", () => {
 		);
 	});

+	// =========================================================================
+	// Cloudflare Workers AI
+	// =========================================================================
+
+	describe.skipIf(!process.env.CLOUDFLARE_API_KEY || !process.env.CLOUDFLARE_ACCOUNT_ID)(
+		"Cloudflare Workers AI",
+		() => {
+			it(
+				"@cf/moonshotai/kimi-k2.6 - should return totalTokens equal to sum of components",
+				{ retry: 3, timeout: 60000 },
+				async () => {
+					const llm = getModel("cloudflare-workers-ai", "@cf/moonshotai/kimi-k2.6");
+
+					console.log(`\nCloudflare Workers AI / ${llm.id}:`);
+					const { first, second } = await testTotalTokensWithCache(llm, {
+						apiKey: process.env.CLOUDFLARE_API_KEY,
+					});
+
+					logUsage("First request", first);
+					logUsage("Second request", second);
+
+					assertTotalTokensEqualsComponents(first);
+					assertTotalTokensEqualsComponents(second);
+				},
+			);
+		},
+	);
+
 	// =========================================================================
 	// Hugging Face
 	// =========================================================================
--- a/packages/ai/test/unicode-surrogate.test.ts
+++ b/packages/ai/test/unicode-surrogate.test.ts
@@ -611,6 +611,29 @@ describe("AI Providers Unicode Surrogate Pair Tests", () => {
 		});
 	});

+	describe.skipIf(!process.env.CLOUDFLARE_API_KEY || !process.env.CLOUDFLARE_ACCOUNT_ID)(
+		"Cloudflare Workers AI Provider Unicode Handling",
+		() => {
+			const llm = getModel("cloudflare-workers-ai", "@cf/moonshotai/kimi-k2.6");
+
+			it("should handle emoji in tool results", { retry: 3, timeout: 30000 }, async () => {
+				await testEmojiInToolResults(llm);
+			});
+
+			it("should handle real-world LinkedIn comment data with emoji", { retry: 3, timeout: 30000 }, async () => {
+				await testRealWorldLinkedInData(llm);
+			});
+
+			it(
+				"should handle unpaired high surrogate (0xD83D) in tool results",
+				{ retry: 3, timeout: 30000 },
+				async () => {
+					await testUnpairedHighSurrogate(llm);
+				},
+			);
+		},
+	);
+
 	describe.skipIf(!process.env.HF_TOKEN)("Hugging Face Provider Unicode Handling", () => {
 		const llm = getModel("huggingface", "moonshotai/Kimi-K2.5");

--- a/packages/coding-agent/README.md
+++ b/packages/coding-agent/README.md
@@ -114,6 +114,7 @@ For each built-in provider, pi maintains a list of tool-capable models, updated
 - Mistral
 - Groq
 - Cerebras
+- Cloudflare Workers AI
 - xAI
 - OpenRouter
 - Vercel AI Gateway
--- a/packages/coding-agent/docs/providers.md
+++ b/packages/coding-agent/docs/providers.md
@@ -61,6 +61,7 @@ pi
 | Mistral | `MISTRAL_API_KEY` | `mistral` |
 | Groq | `GROQ_API_KEY` | `groq` |
 | Cerebras | `CEREBRAS_API_KEY` | `cerebras` |
+| Cloudflare Workers AI | `CLOUDFLARE_API_KEY` (+ `CLOUDFLARE_ACCOUNT_ID`) | `cloudflare-workers-ai` |
 | xAI | `XAI_API_KEY` | `xai` |
 | OpenRouter | `OPENROUTER_API_KEY` | `openrouter` |
 | Vercel AI Gateway | `AI_GATEWAY_API_KEY` | `vercel-ai-gateway` |
@@ -172,6 +173,18 @@ export AWS_BEDROCK_SKIP_AUTH=1
 export AWS_BEDROCK_FORCE_HTTP1=1
 ```

+### Cloudflare Workers AI
+
+`CLOUDFLARE_API_KEY` can be set via `/login`. `CLOUDFLARE_ACCOUNT_ID` must be set as an environment variable.
+
+```bash
+export CLOUDFLARE_API_KEY=...           # or use /login
+export CLOUDFLARE_ACCOUNT_ID=...
+pi --provider cloudflare-workers-ai --model "@cf/moonshotai/kimi-k2.6"
+```
+
+Pi automatically sets `x-session-affinity` for [prefix caching](https://developers.cloudflare.com/workers-ai/features/prompt-caching/) discounts.
+
 ### Google Vertex AI

 Uses Application Default Credentials:
--- a/packages/coding-agent/src/cli/args.ts
+++ b/packages/coding-agent/src/cli/args.ts
@@ -316,6 +316,8 @@ ${chalk.bold("Environment Variables:")}
  MINIMAX_API_KEY                  - MiniMax API key
  OPENCODE_API_KEY                 - OpenCode Zen/OpenCode Go API key
  KIMI_API_KEY                     - Kimi For Coding API key
+  CLOUDFLARE_API_KEY               - Cloudflare API token (Workers AI)
+  CLOUDFLARE_ACCOUNT_ID            - Cloudflare account id (required for Workers AI)
  AWS_PROFILE                      - AWS profile for Amazon Bedrock
  AWS_ACCESS_KEY_ID                - AWS access key for Amazon Bedrock
  AWS_SECRET_ACCESS_KEY            - AWS secret key for Amazon Bedrock
--- a/packages/coding-agent/src/core/model-resolver.ts
+++ b/packages/coding-agent/src/core/model-resolver.ts
@@ -37,6 +37,7 @@ export const defaultModelPerProvider: Record<KnownProvider, string> = {
 	opencode: "kimi-k2.6",
 	"opencode-go": "kimi-k2.6",
 	"kimi-coding": "kimi-for-coding",
+	"cloudflare-workers-ai": "@cf/moonshotai/kimi-k2.6",
 };

 export interface ScopedModel {
--- a/packages/coding-agent/src/modes/interactive/interactive-mode.ts
+++ b/packages/coding-agent/src/modes/interactive/interactive-mode.ts
@@ -182,6 +182,7 @@ const API_KEY_LOGIN_PROVIDERS: Record<string, string> = {
 	[BEDROCK_PROVIDER_ID]: "Amazon Bedrock",
 	"azure-openai-responses": "Azure OpenAI Responses",
 	cerebras: "Cerebras",
+	"cloudflare-workers-ai": "Cloudflare Workers AI",
 	deepseek: "DeepSeek",
 	fireworks: "Fireworks",
 	google: "Google Gemini",