@poolzin/pool-bot 2026.2.24 → 2026.2.26

package/dist/agents/models-config.providers.js

@@ -1,68 +1,68 @@
+import { createSubsystemLogger } from "../logging/subsystem.js";
 import { DEFAULT_COPILOT_API_BASE_URL, resolveCopilotApiToken, } from "../providers/github-copilot-token.js";
 import { ensureAuthProfileStore, listProfilesForProvider } from "./auth-profiles.js";
-import { resolveAwsSdkEnvVarName, resolveEnvApiKey } from "./model-auth.js";
 import { discoverBedrockModels } from "./bedrock-discovery.js";
+import { buildBytePlusModelDefinition, BYTEPLUS_BASE_URL, BYTEPLUS_MODEL_CATALOG, BYTEPLUS_CODING_BASE_URL, BYTEPLUS_CODING_MODEL_CATALOG, } from "./byteplus-models.js";
 import { buildCloudflareAiGatewayModelDefinition, resolveCloudflareAiGatewayBaseUrl, } from "./cloudflare-ai-gateway.js";
+import { buildDoubaoModelDefinition, DOUBAO_BASE_URL, DOUBAO_MODEL_CATALOG, DOUBAO_CODING_BASE_URL, DOUBAO_CODING_MODEL_CATALOG, } from "./doubao-models.js";
+import { discoverHuggingfaceModels, HUGGINGFACE_BASE_URL, HUGGINGFACE_MODEL_CATALOG, buildHuggingfaceModelDefinition, } from "./huggingface-models.js";
+import { resolveAwsSdkEnvVarName, resolveEnvApiKey } from "./model-auth.js";
+import { OLLAMA_NATIVE_BASE_URL } from "./ollama-stream.js";
 import { buildSyntheticModelDefinition, SYNTHETIC_BASE_URL, SYNTHETIC_MODEL_CATALOG, } from "./synthetic-models.js";
 import { TOGETHER_BASE_URL, TOGETHER_MODEL_CATALOG, buildTogetherModelDefinition, } from "./together-models.js";
-import { discoverNvidiaModels, NVIDIA_BASE_URL } from "./nvidia-models.js";
 import { discoverVeniceModels, VENICE_BASE_URL } from "./venice-models.js";
-const MINIMAX_API_BASE_URL = "https://api.minimax.chat/v1";
 const MINIMAX_PORTAL_BASE_URL = "https://api.minimax.io/anthropic";
 const MINIMAX_DEFAULT_MODEL_ID = "MiniMax-M2.1";
 const MINIMAX_DEFAULT_VISION_MODEL_ID = "MiniMax-VL-01";
 const MINIMAX_DEFAULT_CONTEXT_WINDOW = 200000;
 const MINIMAX_DEFAULT_MAX_TOKENS = 8192;
 const MINIMAX_OAUTH_PLACEHOLDER = "minimax-oauth";
-// Pricing: MiniMax doesn't publish public rates. Override in models.json for accurate costs.
+// Pricing per 1M tokens (USD) — https://platform.minimaxi.com/document/Price
 const MINIMAX_API_COST = {
-    input: 15,
-    output: 60,
-    cacheRead: 2,
-    cacheWrite: 10,
+    input: 0.3,
+    output: 1.2,
+    cacheRead: 0.03,
+    cacheWrite: 0.12,
 };
-const MOONSHOT_BASE_URL = "https://api.moonshot.ai/v1";
-const MOONSHOT_DEFAULT_MODEL_ID = "kimi-k2.5";
-const MOONSHOT_DEFAULT_CONTEXT_WINDOW = 256000;
-const MOONSHOT_DEFAULT_MAX_TOKENS = 8192;
-const MOONSHOT_DEFAULT_COST = {
+function buildMinimaxModel(params) {
+    return {
+        id: params.id,
+        name: params.name,
+        reasoning: params.reasoning,
+        input: params.input,
+        cost: MINIMAX_API_COST,
+        contextWindow: MINIMAX_DEFAULT_CONTEXT_WINDOW,
+        maxTokens: MINIMAX_DEFAULT_MAX_TOKENS,
+    };
+}
+function buildMinimaxTextModel(params) {
+    return buildMinimaxModel({ ...params, input: ["text"] });
+}
+const XIAOMI_BASE_URL = "https://api.xiaomimimo.com/anthropic";
+export const XIAOMI_DEFAULT_MODEL_ID = "mimo-v2-flash";
+const XIAOMI_DEFAULT_CONTEXT_WINDOW = 262144;
+const XIAOMI_DEFAULT_MAX_TOKENS = 8192;
+const XIAOMI_DEFAULT_COST = {
     input: 0,
     output: 0,
     cacheRead: 0,
     cacheWrite: 0,
 };
-// Z.AI (ZhipuAI / GLM) — OpenAI-compatible API
-// Docs: https://docs.z.ai/ | Base: https://open.bigmodel.cn/api/paas/v4/
-const ZAI_BASE_URL = "https://open.bigmodel.cn/api/paas/v4";
-const ZAI_DEFAULT_MODEL_ID = "GLM-4.7-Flash";
-const ZAI_DEFAULT_CONTEXT_WINDOW = 128000;
-const ZAI_DEFAULT_MAX_TOKENS = 8192;
-// GLM-4.7-Flash and GLM-4.5-Flash are completely free
-const ZAI_FREE_COST = {
+const MOONSHOT_BASE_URL = "https://api.moonshot.ai/v1";
+const MOONSHOT_DEFAULT_MODEL_ID = "kimi-k2.5";
+const MOONSHOT_DEFAULT_CONTEXT_WINDOW = 256000;
+const MOONSHOT_DEFAULT_MAX_TOKENS = 8192;
+const MOONSHOT_DEFAULT_COST = {
     input: 0,
     output: 0,
     cacheRead: 0,
     cacheWrite: 0,
 };
-// GLM-4.7-FlashX: $0.07/$0.4 per 1M tokens
-const ZAI_FLASHX_COST = {
-    input: 0.07,
-    output: 0.4,
-    cacheRead: 0,
-    cacheWrite: 0,
-};
-// GLM-4.7: $0.6/$2.2 per 1M tokens
-const ZAI_PREMIUM_COST = {
-    input: 0.6,
-    output: 2.2,
-    cacheRead: 0,
-    cacheWrite: 0,
-};
-const XIAOMI_BASE_URL = "https://api.xiaomimimo.com/anthropic";
-export const XIAOMI_DEFAULT_MODEL_ID = "mimo-v2-flash";
-const XIAOMI_DEFAULT_CONTEXT_WINDOW = 262144;
-const XIAOMI_DEFAULT_MAX_TOKENS = 8192;
-const XIAOMI_DEFAULT_COST = {
+const KIMI_CODING_BASE_URL = "https://api.kimi.com/coding/";
+const KIMI_CODING_DEFAULT_MODEL_ID = "k2p5";
+const KIMI_CODING_DEFAULT_CONTEXT_WINDOW = 262144;
+const KIMI_CODING_DEFAULT_MAX_TOKENS = 32768;
+const KIMI_CODING_DEFAULT_COST = {
     input: 0,
     output: 0,
     cacheRead: 0,
@@ -78,8 +78,8 @@ const QWEN_PORTAL_DEFAULT_COST = {
     cacheRead: 0,
     cacheWrite: 0,
 };
-const OLLAMA_BASE_URL = "http://127.0.0.1:11434/v1";
-const OLLAMA_API_BASE_URL = "http://127.0.0.1:11434";
+const OLLAMA_BASE_URL = OLLAMA_NATIVE_BASE_URL;
+const OLLAMA_API_BASE_URL = OLLAMA_BASE_URL;
 const OLLAMA_DEFAULT_CONTEXT_WINDOW = 128000;
 const OLLAMA_DEFAULT_MAX_TOKENS = 8192;
 const OLLAMA_DEFAULT_COST = {
@@ -88,6 +88,15 @@ const OLLAMA_DEFAULT_COST = {
     cacheRead: 0,
     cacheWrite: 0,
 };
+const VLLM_BASE_URL = "http://127.0.0.1:8000/v1";
+const VLLM_DEFAULT_CONTEXT_WINDOW = 128000;
+const VLLM_DEFAULT_MAX_TOKENS = 8192;
+const VLLM_DEFAULT_COST = {
+    input: 0,
+    output: 0,
+    cacheRead: 0,
+    cacheWrite: 0,
+};
 export const QIANFAN_BASE_URL = "https://qianfan.baidubce.com/v2";
 export const QIANFAN_DEFAULT_MODEL_ID = "deepseek-v3.2";
 const QIANFAN_DEFAULT_CONTEXT_WINDOW = 98304;
@@ -98,22 +107,50 @@ const QIANFAN_DEFAULT_COST = {
     cacheRead: 0,
     cacheWrite: 0,
 };
-async function discoverOllamaModels() {
+const NVIDIA_BASE_URL = "https://integrate.api.nvidia.com/v1";
+const NVIDIA_DEFAULT_MODEL_ID = "nvidia/llama-3.1-nemotron-70b-instruct";
+const NVIDIA_DEFAULT_CONTEXT_WINDOW = 131072;
+const NVIDIA_DEFAULT_MAX_TOKENS = 4096;
+const NVIDIA_DEFAULT_COST = {
+    input: 0,
+    output: 0,
+    cacheRead: 0,
+    cacheWrite: 0,
+};
+const log = createSubsystemLogger("agents/model-providers");
+/**
+ * Derive the Ollama native API base URL from a configured base URL.
+ *
+ * Users typically configure `baseUrl` with a `/v1` suffix (e.g.
+ * `http://192.168.20.14:11434/v1`) for the OpenAI-compatible endpoint.
+ * The native Ollama API lives at the root (e.g. `/api/tags`), so we
+ * strip the `/v1` suffix when present.
+ */
+export function resolveOllamaApiBase(configuredBaseUrl) {
+    if (!configuredBaseUrl) {
+        return OLLAMA_API_BASE_URL;
+    }
+    // Strip trailing slash, then strip /v1 suffix if present
+    const trimmed = configuredBaseUrl.replace(/\/+$/, "");
+    return trimmed.replace(/\/v1$/i, "");
+}
+async function discoverOllamaModels(baseUrl) {
     // Skip Ollama discovery in test environments
     if (process.env.VITEST || process.env.NODE_ENV === "test") {
         return [];
     }
     try {
-        const response = await fetch(`${OLLAMA_API_BASE_URL}/api/tags`, {
+        const apiBase = resolveOllamaApiBase(baseUrl);
+        const response = await fetch(`${apiBase}/api/tags`, {
             signal: AbortSignal.timeout(5000),
         });
         if (!response.ok) {
-            console.warn(`Failed to discover Ollama models: ${response.status}`);
+            log.warn(`Failed to discover Ollama models: ${response.status}`);
             return [];
         }
         const data = (await response.json());
         if (!data.models || data.models.length === 0) {
-            console.warn("No Ollama models found on local instance");
+            log.warn("No Ollama models found on local instance");
             return [];
         }
         return data.models.map((model) => {
@@ -127,16 +164,57 @@ async function discoverOllamaModels() {
                 cost: OLLAMA_DEFAULT_COST,
                 contextWindow: OLLAMA_DEFAULT_CONTEXT_WINDOW,
                 maxTokens: OLLAMA_DEFAULT_MAX_TOKENS,
-                // Disable streaming by default for Ollama to avoid SDK issue #1205
-                // See: https://github.com/badlogic/pi-mono/issues/1205
-                params: {
-                    streaming: false,
-                },
             };
         });
     }
     catch (error) {
-        console.warn(`Failed to discover Ollama models: ${String(error)}`);
+        log.warn(`Failed to discover Ollama models: ${String(error)}`);
+        return [];
+    }
+}
+async function discoverVllmModels(baseUrl, apiKey) {
+    // Skip vLLM discovery in test environments
+    if (process.env.VITEST || process.env.NODE_ENV === "test") {
+        return [];
+    }
+    const trimmedBaseUrl = baseUrl.trim().replace(/\/+$/, "");
+    const url = `${trimmedBaseUrl}/models`;
+    try {
+        const trimmedApiKey = apiKey?.trim();
+        const response = await fetch(url, {
+            headers: trimmedApiKey ? { Authorization: `Bearer ${trimmedApiKey}` } : undefined,
+            signal: AbortSignal.timeout(5000),
+        });
+        if (!response.ok) {
+            log.warn(`Failed to discover vLLM models: ${response.status}`);
+            return [];
+        }
+        const data = (await response.json());
+        const models = data.data ?? [];
+        if (models.length === 0) {
+            log.warn("No vLLM models found on local instance");
+            return [];
+        }
+        return models
+            .map((m) => ({ id: typeof m.id === "string" ? m.id.trim() : "" }))
+            .filter((m) => Boolean(m.id))
+            .map((m) => {
+            const modelId = m.id;
+            const lower = modelId.toLowerCase();
+            const isReasoning = lower.includes("r1") || lower.includes("reasoning") || lower.includes("think");
+            return {
+                id: modelId,
+                name: modelId,
+                reasoning: isReasoning,
+                input: ["text"],
+                cost: VLLM_DEFAULT_COST,
+                contextWindow: VLLM_DEFAULT_CONTEXT_WINDOW,
+                maxTokens: VLLM_DEFAULT_MAX_TOKENS,
+            };
+        });
+    }
+    catch (error) {
+        log.warn(`Failed to discover vLLM models: ${String(error)}`);
         return [];
     }
 }
@@ -147,8 +225,9 @@ function normalizeApiKeyConfig(value) {
 }
 function resolveEnvApiKeyVarName(provider) {
     const resolved = resolveEnvApiKey(provider);
-    if (!resolved)
+    if (!resolved) {
         return undefined;
+    }
     const match = /^(?:env: |shell env: )([A-Z0-9_]+)$/.exec(resolved.source);
     return match ? match[1] : undefined;
 }
@@ -159,28 +238,34 @@ function resolveApiKeyFromProfiles(params) {
     const ids = listProfilesForProvider(params.store, params.provider);
     for (const id of ids) {
         const cred = params.store.profiles[id];
-        if (!cred)
+        if (!cred) {
             continue;
-        if (cred.type === "api_key")
+        }
+        if (cred.type === "api_key") {
             return cred.key;
-        if (cred.type === "token")
+        }
+        if (cred.type === "token") {
             return cred.token;
+        }
     }
     return undefined;
 }
 export function normalizeGoogleModelId(id) {
-    if (id === "gemini-3-pro")
+    if (id === "gemini-3-pro") {
         return "gemini-3-pro-preview";
-    if (id === "gemini-3-flash")
+    }
+    if (id === "gemini-3-flash") {
         return "gemini-3-flash-preview";
+    }
     return id;
 }
 function normalizeGoogleProvider(provider) {
     let mutated = false;
     const models = provider.models.map((model) => {
         const nextId = normalizeGoogleModelId(model.id);
-        if (nextId === model.id)
+        if (nextId === model.id) {
             return model;
+        }
         mutated = true;
         return { ...model, id: nextId };
     });
@@ -188,8 +273,9 @@ function normalizeGoogleProvider(provider) {
 }
 export function normalizeProviders(params) {
     const { providers } = params;
-    if (!providers)
+    if (!providers) {
         return providers;
+    }
     const authStore = ensureAuthProfileStore(params.agentDir, {
         allowKeychainPrompt: false,
     });
@@ -232,8 +318,9 @@ export function normalizeProviders(params) {
         }
         if (normalizedKey === "google") {
             const googleNormalized = normalizeGoogleProvider(normalizedProvider);
-            if (googleNormalized !== normalizedProvider)
+            if (googleNormalized !== normalizedProvider) {
                 mutated = true;
+            }
             normalizedProvider = googleNormalized;
         }
         next[key] = normalizedProvider;
@@ -242,27 +329,53 @@ export function normalizeProviders(params) {
 }
 function buildMinimaxProvider() {
     return {
-        baseUrl: MINIMAX_API_BASE_URL,
-        api: "openai-completions",
+        baseUrl: MINIMAX_PORTAL_BASE_URL,
+        api: "anthropic-messages",
         models: [
-            {
+            buildMinimaxTextModel({
                 id: MINIMAX_DEFAULT_MODEL_ID,
                 name: "MiniMax M2.1",
                 reasoning: false,
-                input: ["text"],
-                cost: MINIMAX_API_COST,
-                contextWindow: MINIMAX_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: MINIMAX_DEFAULT_MAX_TOKENS,
-            },
-            {
+            }),
+            buildMinimaxTextModel({
+                id: "MiniMax-M2.1-lightning",
+                name: "MiniMax M2.1 Lightning",
+                reasoning: false,
+            }),
+            buildMinimaxModel({
                 id: MINIMAX_DEFAULT_VISION_MODEL_ID,
                 name: "MiniMax VL 01",
                 reasoning: false,
                 input: ["text", "image"],
-                cost: MINIMAX_API_COST,
-                contextWindow: MINIMAX_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: MINIMAX_DEFAULT_MAX_TOKENS,
-            },
+            }),
+            buildMinimaxTextModel({
+                id: "MiniMax-M2.5",
+                name: "MiniMax M2.5",
+                reasoning: true,
+            }),
+            buildMinimaxTextModel({
+                id: "MiniMax-M2.5-Lightning",
+                name: "MiniMax M2.5 Lightning",
+                reasoning: true,
+            }),
+        ],
+    };
+}
+function buildMinimaxPortalProvider() {
+    return {
+        baseUrl: MINIMAX_PORTAL_BASE_URL,
+        api: "anthropic-messages",
+        models: [
+            buildMinimaxTextModel({
+                id: MINIMAX_DEFAULT_MODEL_ID,
+                name: "MiniMax M2.1",
+                reasoning: false,
+            }),
+            buildMinimaxTextModel({
+                id: "MiniMax-M2.5",
+                name: "MiniMax M2.5",
+                reasoning: true,
+            }),
         ],
     };
 }
@@ -283,55 +396,19 @@ function buildMoonshotProvider() {
         ],
     };
 }
-function buildZaiProvider() {
+export function buildKimiCodingProvider() {
     return {
-        baseUrl: ZAI_BASE_URL,
-        api: "openai-completions",
+        baseUrl: KIMI_CODING_BASE_URL,
+        api: "anthropic-messages",
         models: [
             {
-                id: ZAI_DEFAULT_MODEL_ID,
-                name: "GLM 4.7 Flash",
-                reasoning: false,
-                input: ["text"],
-                cost: ZAI_FREE_COST,
-                contextWindow: ZAI_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: ZAI_DEFAULT_MAX_TOKENS,
-            },
-            {
-                id: "GLM-4.5-Flash",
-                name: "GLM 4.5 Flash",
-                reasoning: false,
-                input: ["text"],
-                cost: ZAI_FREE_COST,
-                contextWindow: ZAI_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: ZAI_DEFAULT_MAX_TOKENS,
-            },
-            {
-                id: "GLM-4.6V-Flash",
-                name: "GLM 4.6V Flash (Vision)",
-                reasoning: false,
+                id: KIMI_CODING_DEFAULT_MODEL_ID,
+                name: "Kimi for Coding",
+                reasoning: true,
                 input: ["text", "image"],
-                cost: ZAI_FREE_COST,
-                contextWindow: ZAI_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: ZAI_DEFAULT_MAX_TOKENS,
-            },
-            {
-                id: "GLM-4.7-FlashX",
-                name: "GLM 4.7 FlashX",
-                reasoning: false,
-                input: ["text"],
-                cost: ZAI_FLASHX_COST,
-                contextWindow: ZAI_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: ZAI_DEFAULT_MAX_TOKENS,
-            },
-            {
-                id: "GLM-4.7",
-                name: "GLM 4.7",
-                reasoning: false,
-                input: ["text"],
-                cost: ZAI_PREMIUM_COST,
-                contextWindow: ZAI_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: ZAI_DEFAULT_MAX_TOKENS,
+                cost: KIMI_CODING_DEFAULT_COST,
+                contextWindow: KIMI_CODING_DEFAULT_CONTEXT_WINDOW,
+                maxTokens: KIMI_CODING_DEFAULT_MAX_TOKENS,
             },
         ],
     };
@@ -369,37 +446,32 @@ function buildSyntheticProvider() {
         models: SYNTHETIC_MODEL_CATALOG.map(buildSyntheticModelDefinition),
     };
 }
-async function buildVeniceProvider() {
-    const models = await discoverVeniceModels();
+function buildDoubaoProvider() {
     return {
-        baseUrl: VENICE_BASE_URL,
+        baseUrl: DOUBAO_BASE_URL,
         api: "openai-completions",
-        models,
+        models: DOUBAO_MODEL_CATALOG.map(buildDoubaoModelDefinition),
     };
 }
-async function buildOllamaProvider() {
-    const models = await discoverOllamaModels();
+function buildDoubaoCodingProvider() {
     return {
-        baseUrl: OLLAMA_BASE_URL,
+        baseUrl: DOUBAO_CODING_BASE_URL,
         api: "openai-completions",
-        models,
+        models: DOUBAO_CODING_MODEL_CATALOG.map(buildDoubaoModelDefinition),
     };
 }
-function buildMinimaxPortalProvider() {
+function buildBytePlusProvider() {
     return {
-        baseUrl: MINIMAX_PORTAL_BASE_URL,
-        api: "anthropic-messages",
-        models: [
-            {
-                id: MINIMAX_DEFAULT_MODEL_ID,
-                name: "MiniMax M2.1",
-                reasoning: false,
-                input: ["text"],
-                cost: MINIMAX_API_COST,
-                contextWindow: MINIMAX_DEFAULT_CONTEXT_WINDOW,
-                maxTokens: MINIMAX_DEFAULT_MAX_TOKENS,
-            },
-        ],
+        baseUrl: BYTEPLUS_BASE_URL,
+        api: "openai-completions",
+        models: BYTEPLUS_MODEL_CATALOG.map(buildBytePlusModelDefinition),
+    };
+}
+function buildBytePlusCodingProvider() {
+    return {
+        baseUrl: BYTEPLUS_CODING_BASE_URL,
+        api: "openai-completions",
+        models: BYTEPLUS_CODING_MODEL_CATALOG.map(buildBytePlusModelDefinition),
     };
 }
 export function buildXiaomiProvider() {
@@ -419,6 +491,38 @@ export function buildXiaomiProvider() {
         ],
     };
 }
+async function buildVeniceProvider() {
+    const models = await discoverVeniceModels();
+    return {
+        baseUrl: VENICE_BASE_URL,
+        api: "openai-completions",
+        models,
+    };
+}
+async function buildOllamaProvider(configuredBaseUrl) {
+    const models = await discoverOllamaModels(configuredBaseUrl);
+    return {
+        baseUrl: resolveOllamaApiBase(configuredBaseUrl),
+        api: "ollama",
+        models,
+    };
+}
+async function buildHuggingfaceProvider(apiKey) {
+    // Resolve env var name to value for discovery (GET /v1/models requires Bearer token).
+    const resolvedSecret = apiKey?.trim() !== ""
+        ? /^[A-Z][A-Z0-9_]*$/.test(apiKey.trim())
+            ? (process.env[apiKey.trim()] ?? "").trim()
+            : apiKey.trim()
+        : "";
+    const models = resolvedSecret !== ""
+        ? await discoverHuggingfaceModels(resolvedSecret)
+        : HUGGINGFACE_MODEL_CATALOG.map(buildHuggingfaceModelDefinition);
+    return {
+        baseUrl: HUGGINGFACE_BASE_URL,
+        api: "openai-completions",
+        models,
+    };
+}
 function buildTogetherProvider() {
     return {
         baseUrl: TOGETHER_BASE_URL,
@@ -426,10 +530,11 @@ function buildTogetherProvider() {
         models: TOGETHER_MODEL_CATALOG.map(buildTogetherModelDefinition),
     };
 }
-async function buildNvidiaProvider(apiKey) {
-    const models = await discoverNvidiaModels(apiKey);
+async function buildVllmProvider(params) {
+    const baseUrl = (params?.baseUrl?.trim() || VLLM_BASE_URL).replace(/\/+$/, "");
+    const models = await discoverVllmModels(baseUrl, params?.apiKey);
     return {
-        baseUrl: NVIDIA_BASE_URL,
+        baseUrl,
         api: "openai-completions",
         models,
     };
@@ -460,6 +565,41 @@ export function buildQianfanProvider() {
         ],
     };
 }
+export function buildNvidiaProvider() {
+    return {
+        baseUrl: NVIDIA_BASE_URL,
+        api: "openai-completions",
+        models: [
+            {
+                id: NVIDIA_DEFAULT_MODEL_ID,
+                name: "NVIDIA Llama 3.1 Nemotron 70B Instruct",
+                reasoning: false,
+                input: ["text"],
+                cost: NVIDIA_DEFAULT_COST,
+                contextWindow: NVIDIA_DEFAULT_CONTEXT_WINDOW,
+                maxTokens: NVIDIA_DEFAULT_MAX_TOKENS,
+            },
+            {
+                id: "meta/llama-3.3-70b-instruct",
+                name: "Meta Llama 3.3 70B Instruct",
+                reasoning: false,
+                input: ["text"],
+                cost: NVIDIA_DEFAULT_COST,
+                contextWindow: 131072,
+                maxTokens: 4096,
+            },
+            {
+                id: "nvidia/mistral-nemo-minitron-8b-8k-instruct",
+                name: "NVIDIA Mistral NeMo Minitron 8B Instruct",
+                reasoning: false,
+                input: ["text"],
+                cost: NVIDIA_DEFAULT_COST,
+                contextWindow: 8192,
+                maxTokens: 2048,
+            },
+        ],
+    };
+}
 export async function resolveImplicitProviders(params) {
     const providers = {};
     const authStore = ensureAuthProfileStore(params.agentDir, {
@@ -482,10 +622,10 @@ export async function resolveImplicitProviders(params) {
     if (moonshotKey) {
         providers.moonshot = { ...buildMoonshotProvider(), apiKey: moonshotKey };
     }
-    const zaiKey = resolveEnvApiKeyVarName("zai") ??
-        resolveApiKeyFromProfiles({ provider: "zai", store: authStore });
-    if (zaiKey) {
-        providers.zai = { ...buildZaiProvider(), apiKey: zaiKey };
+    const kimiCodingKey = resolveEnvApiKeyVarName("kimi-coding") ??
+        resolveApiKeyFromProfiles({ provider: "kimi-coding", store: authStore });
+    if (kimiCodingKey) {
+        providers["kimi-coding"] = { ...buildKimiCodingProvider(), apiKey: kimiCodingKey };
     }
     const syntheticKey = resolveEnvApiKeyVarName("synthetic") ??
         resolveApiKeyFromProfiles({ provider: "synthetic", store: authStore });
@@ -504,6 +644,24 @@ export async function resolveImplicitProviders(params) {
             apiKey: QWEN_PORTAL_OAUTH_PLACEHOLDER,
         };
     }
+    const volcengineKey = resolveEnvApiKeyVarName("volcengine") ??
+        resolveApiKeyFromProfiles({ provider: "volcengine", store: authStore });
+    if (volcengineKey) {
+        providers.volcengine = { ...buildDoubaoProvider(), apiKey: volcengineKey };
+        providers["volcengine-plan"] = {
+            ...buildDoubaoCodingProvider(),
+            apiKey: volcengineKey,
+        };
+    }
+    const byteplusKey = resolveEnvApiKeyVarName("byteplus") ??
+        resolveApiKeyFromProfiles({ provider: "byteplus", store: authStore });
+    if (byteplusKey) {
+        providers.byteplus = { ...buildBytePlusProvider(), apiKey: byteplusKey };
+        providers["byteplus-plan"] = {
+            ...buildBytePlusCodingProvider(),
+            apiKey: byteplusKey,
+        };
+    }
     const xiaomiKey = resolveEnvApiKeyVarName("xiaomi") ??
         resolveApiKeyFromProfiles({ provider: "xiaomi", store: authStore });
     if (xiaomiKey) {
@@ -512,18 +670,22 @@ export async function resolveImplicitProviders(params) {
     const cloudflareProfiles = listProfilesForProvider(authStore, "cloudflare-ai-gateway");
     for (const profileId of cloudflareProfiles) {
         const cred = authStore.profiles[profileId];
-        if (cred?.type !== "api_key")
+        if (cred?.type !== "api_key") {
             continue;
+        }
         const accountId = cred.metadata?.accountId?.trim();
         const gatewayId = cred.metadata?.gatewayId?.trim();
-        if (!accountId || !gatewayId)
+        if (!accountId || !gatewayId) {
             continue;
+        }
         const baseUrl = resolveCloudflareAiGatewayBaseUrl({ accountId, gatewayId });
-        if (!baseUrl)
+        if (!baseUrl) {
             continue;
+        }
         const apiKey = resolveEnvApiKeyVarName("cloudflare-ai-gateway") ?? cred.key?.trim() ?? "";
-        if (!apiKey)
+        if (!apiKey) {
             continue;
+        }
         providers["cloudflare-ai-gateway"] = {
             baseUrl,
             api: "anthropic-messages",
@@ -532,11 +694,30 @@ export async function resolveImplicitProviders(params) {
         };
         break;
     }
-    // Ollama provider - only add if explicitly configured
+    // Ollama provider - only add if explicitly configured.
+    // Use the user's configured baseUrl (from explicit providers) for model
+    // discovery so that remote / non-default Ollama instances are reachable.
     const ollamaKey = resolveEnvApiKeyVarName("ollama") ??
         resolveApiKeyFromProfiles({ provider: "ollama", store: authStore });
     if (ollamaKey) {
-        providers.ollama = { ...(await buildOllamaProvider()), apiKey: ollamaKey };
+        const ollamaBaseUrl = params.explicitProviders?.ollama?.baseUrl;
+        providers.ollama = { ...(await buildOllamaProvider(ollamaBaseUrl)), apiKey: ollamaKey };
+    }
+    // vLLM provider - OpenAI-compatible local server (opt-in via env/profile).
+    // If explicitly configured, keep user-defined models/settings as-is.
+    if (!params.explicitProviders?.vllm) {
+        const vllmEnvVar = resolveEnvApiKeyVarName("vllm");
+        const vllmProfileKey = resolveApiKeyFromProfiles({ provider: "vllm", store: authStore });
+        const vllmKey = vllmEnvVar ?? vllmProfileKey;
+        if (vllmKey) {
+            const discoveryApiKey = vllmEnvVar
+                ? (process.env[vllmEnvVar]?.trim() ?? "")
+                : (vllmProfileKey ?? "");
+            providers.vllm = {
+                ...(await buildVllmProvider({ apiKey: discoveryApiKey || undefined })),
+                apiKey: vllmKey,
+            };
+        }
     }
     const togetherKey = resolveEnvApiKeyVarName("together") ??
         resolveApiKeyFromProfiles({ provider: "together", store: authStore });
@@ -546,6 +727,15 @@ export async function resolveImplicitProviders(params) {
             apiKey: togetherKey,
         };
     }
+    const huggingfaceKey = resolveEnvApiKeyVarName("huggingface") ??
+        resolveApiKeyFromProfiles({ provider: "huggingface", store: authStore });
+    if (huggingfaceKey) {
+        const hfProvider = await buildHuggingfaceProvider(huggingfaceKey);
+        providers.huggingface = {
+            ...hfProvider,
+            apiKey: huggingfaceKey,
+        };
+    }
     const qianfanKey = resolveEnvApiKeyVarName("qianfan") ??
         resolveApiKeyFromProfiles({ provider: "qianfan", store: authStore });
     if (qianfanKey) {
@@ -554,18 +744,21 @@ export async function resolveImplicitProviders(params) {
     const nvidiaKey = resolveEnvApiKeyVarName("nvidia") ??
         resolveApiKeyFromProfiles({ provider: "nvidia", store: authStore });
     if (nvidiaKey) {
-        providers.nvidia = { ...(await buildNvidiaProvider(nvidiaKey)), apiKey: nvidiaKey };
+        providers.nvidia = { ...buildNvidiaProvider(), apiKey: nvidiaKey };
     }
     return providers;
 }
 export async function resolveImplicitCopilotProvider(params) {
     const env = params.env ?? process.env;
-    const authStore = ensureAuthProfileStore(params.agentDir, { allowKeychainPrompt: false });
+    const authStore = ensureAuthProfileStore(params.agentDir, {
+        allowKeychainPrompt: false,
+    });
     const hasProfile = listProfilesForProvider(authStore, "github-copilot").length > 0;
     const envToken = env.COPILOT_GITHUB_TOKEN ?? env.GH_TOKEN ?? env.GITHUB_TOKEN;
     const githubToken = (envToken ?? "").trim();
-    if (!hasProfile && !githubToken)
+    if (!hasProfile && !githubToken) {
         return null;
+    }
     let selectedGithubToken = githubToken;
     if (!selectedGithubToken && hasProfile) {
         // Use the first available profile as a default for discovery (it will be
@@ -591,15 +784,15 @@ export async function resolveImplicitCopilotProvider(params) {
     }
     // pi-coding-agent's ModelRegistry marks a model "available" only if its
     // `AuthStorage` has auth configured for that provider (via auth.json/env/etc).
-    // Our Copilot auth lives in Poolbot's auth-profiles store instead, so we also
+    // Our Copilot auth lives in Pool Bot's auth-profiles store instead, so we also
     // write a runtime-only auth.json entry for pi-coding-agent to pick up.
     //
-    // This is safe because it's (1) within Poolbot's agent dir, (2) contains the
+    // This is safe because it's (1) within Pool Bot's agent dir, (2) contains the
     // GitHub token (not the exchanged Copilot token), and (3) matches existing
     // patterns for OAuth-like providers in pi-coding-agent.
     // Note: we deliberately do not write pi-coding-agent's `auth.json` here.
-    // Poolbot uses its own auth store and exchanges tokens at runtime.
-    // `models list` uses Poolbot's auth heuristics for availability.
+    // Pool Bot uses its own auth store and exchanges tokens at runtime.
+    // `models list` uses Pool Bot's auth heuristics for availability.
     // We intentionally do NOT define custom models for Copilot in models.json.
     // pi-coding-agent treats providers with models as replacements requiring apiKey.
     // We only override baseUrl; the model list comes from pi-ai built-ins.
@@ -613,14 +806,20 @@ export async function resolveImplicitBedrockProvider(params) {
     const discoveryConfig = params.config?.models?.bedrockDiscovery;
     const enabled = discoveryConfig?.enabled;
     const hasAwsCreds = resolveAwsSdkEnvVarName(env) !== undefined;
-    if (enabled === false)
+    if (enabled === false) {
         return null;
-    if (enabled !== true && !hasAwsCreds)
+    }
+    if (enabled !== true && !hasAwsCreds) {
         return null;
+    }
     const region = discoveryConfig?.region ?? env.AWS_REGION ?? env.AWS_DEFAULT_REGION ?? "us-east-1";
-    const models = await discoverBedrockModels({ region, config: discoveryConfig });
-    if (models.length === 0)
+    const models = await discoverBedrockModels({
+        region,
+        config: discoveryConfig,
+    });
+    if (models.length === 0) {
         return null;
+    }
     return {
         baseUrl: `https://bedrock-runtime.${region}.amazonaws.com`,
         api: "bedrock-converse-stream",