Onboard: add Ollama auth flow and improve model defaults

Add Ollama as a auth provider in onboarding with Cloud + Local mode selection, browser-based sign-in via /api/me, smart model suggestions per mode, and graceful fallback when the default model is unavailable. - Extract shared ollama-models.ts - Auto-pull missing models during onboarding - Non-interactive mode support for CI/automation Closes #8239 Closes #3494 Co-Authored-By: Jeffrey Morgan <jmorganca@gmail.com>
2026-03-12 07:20:45 +00:00 · 2026-03-03 17:21:37 -08:00
parent 62d5df28dc
commit d6108a6f72
16 changed files with 1176 additions and 61 deletions
--- a/src/agents/huggingface-models.ts
+++ b/src/agents/huggingface-models.ts
@@ -1,5 +1,6 @@
 import type { ModelDefinitionConfig } from "../config/types.models.js";
 import { createSubsystemLogger } from "../logging/subsystem.js";
+import { isReasoningModelHeuristic } from "./ollama-models.js";

 const log = createSubsystemLogger("huggingface-models");

@@ -125,7 +126,7 @@ export function buildHuggingfaceModelDefinition(
 */
 function inferredMetaFromModelId(id: string): { name: string; reasoning: boolean } {
  const base = id.split("/").pop() ?? id;
-  const reasoning = /r1|reasoning|thinking|reason/i.test(id) || /-\d+[tb]?-thinking/i.test(base);
+  const reasoning = isReasoningModelHeuristic(id);
  const name = base.replace(/-/g, " ").replace(/\b(\w)/g, (c) => c.toUpperCase());
  return { name, reasoning };
 }
--- a/src/agents/models-config.providers.discovery.ts
+++ b/src/agents/models-config.providers.discovery.ts
@@ -9,27 +9,26 @@ import {
  buildHuggingfaceModelDefinition,
 } from "./huggingface-models.js";
 import { discoverKilocodeModels } from "./kilocode-models.js";
-import { OLLAMA_NATIVE_BASE_URL } from "./ollama-stream.js";
+import {
+  OLLAMA_DEFAULT_CONTEXT_WINDOW,
+  OLLAMA_DEFAULT_COST,
+  OLLAMA_DEFAULT_MAX_TOKENS,
+  isReasoningModelHeuristic,
+  resolveOllamaApiBase,
+  type OllamaTagsResponse,
+} from "./ollama-models.js";
 import { discoverVeniceModels, VENICE_BASE_URL } from "./venice-models.js";
 import { discoverVercelAiGatewayModels, VERCEL_AI_GATEWAY_BASE_URL } from "./vercel-ai-gateway.js";

+export { resolveOllamaApiBase } from "./ollama-models.js";
+
 type ModelsConfig = NonNullable<OpenClawConfig["models"]>;
 type ProviderConfig = NonNullable<ModelsConfig["providers"]>[string];

 const log = createSubsystemLogger("agents/model-providers");

-const OLLAMA_BASE_URL = OLLAMA_NATIVE_BASE_URL;
-const OLLAMA_API_BASE_URL = OLLAMA_BASE_URL;
 const OLLAMA_SHOW_CONCURRENCY = 8;
 const OLLAMA_SHOW_MAX_MODELS = 200;
-const OLLAMA_DEFAULT_CONTEXT_WINDOW = 128000;
-const OLLAMA_DEFAULT_MAX_TOKENS = 8192;
-const OLLAMA_DEFAULT_COST = {
-  input: 0,
-  output: 0,
-  cacheRead: 0,
-  cacheWrite: 0,
-};

 const VLLM_BASE_URL = "http://127.0.0.1:8000/v1";
 const VLLM_DEFAULT_CONTEXT_WINDOW = 128000;
@@ -41,44 +40,12 @@ const VLLM_DEFAULT_COST = {
  cacheWrite: 0,
 };

-interface OllamaModel {
-  name: string;
-  modified_at: string;
-  size: number;
-  digest: string;
-  details?: {
-    family?: string;
-    parameter_size?: string;
-  };
-}
-
-interface OllamaTagsResponse {
-  models: OllamaModel[];
-}
-
 type VllmModelsResponse = {
  data?: Array<{
    id?: string;
  }>;
 };

-/**
- * 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?: string): string {
-  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 queryOllamaContextWindow(
  apiBase: string,
  modelName: string,
@@ -147,12 +114,10 @@ async function discoverOllamaModels(
        batch.map(async (model) => {
          const modelId = model.name;
          const contextWindow = await queryOllamaContextWindow(apiBase, modelId);
-          const isReasoning =
-            modelId.toLowerCase().includes("r1") || modelId.toLowerCase().includes("reasoning");
          return {
            id: modelId,
            name: modelId,
-            reasoning: isReasoning,
+            reasoning: isReasoningModelHeuristic(modelId),
            input: ["text"],
            cost: OLLAMA_DEFAULT_COST,
            contextWindow: contextWindow ?? OLLAMA_DEFAULT_CONTEXT_WINDOW,
@@ -204,13 +169,10 @@ async function discoverVllmModels(
      .filter((model) => Boolean(model.id))
      .map((model) => {
        const modelId = model.id;
-        const lower = modelId.toLowerCase();
-        const isReasoning =
-          lower.includes("r1") || lower.includes("reasoning") || lower.includes("think");
        return {
          id: modelId,
          name: modelId,
-          reasoning: isReasoning,
+          reasoning: isReasoningModelHeuristic(modelId),
          input: ["text"],
          cost: VLLM_DEFAULT_COST,
          contextWindow: VLLM_DEFAULT_CONTEXT_WINDOW,
--- a/src/agents/ollama-models.ts
+++ b/src/agents/ollama-models.ts
@@ -0,0 +1,85 @@
+import type { ModelDefinitionConfig } from "../config/types.models.js";
+import { OLLAMA_NATIVE_BASE_URL } from "./ollama-stream.js";
+
+export const OLLAMA_DEFAULT_BASE_URL = OLLAMA_NATIVE_BASE_URL;
+export const OLLAMA_DEFAULT_CONTEXT_WINDOW = 128000;
+export const OLLAMA_DEFAULT_MAX_TOKENS = 8192;
+export const OLLAMA_DEFAULT_COST = {
+  input: 0,
+  output: 0,
+  cacheRead: 0,
+  cacheWrite: 0,
+};
+
+export type OllamaTagModel = {
+  name: string;
+  modified_at?: string;
+  size?: number;
+  digest?: string;
+  remote_host?: string;
+  details?: {
+    family?: string;
+    parameter_size?: string;
+  };
+};
+
+export type OllamaTagsResponse = {
+  models?: OllamaTagModel[];
+};
+
+/**
+ * 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?: string): string {
+  if (!configuredBaseUrl) {
+    return OLLAMA_DEFAULT_BASE_URL;
+  }
+  const trimmed = configuredBaseUrl.replace(/\/+$/, "");
+  return trimmed.replace(/\/v1$/i, "");
+}
+
+/** Heuristic: treat models with "r1", "reasoning", or "think" in the name as reasoning models. */
+export function isReasoningModelHeuristic(modelId: string): boolean {
+  return /r1|reasoning|think|reason/i.test(modelId);
+}
+
+/** Build a ModelDefinitionConfig for an Ollama model with default values. */
+export function buildOllamaModelDefinition(
+  modelId: string,
+  contextWindow?: number,
+): ModelDefinitionConfig {
+  return {
+    id: modelId,
+    name: modelId,
+    reasoning: isReasoningModelHeuristic(modelId),
+    input: ["text"],
+    cost: OLLAMA_DEFAULT_COST,
+    contextWindow: contextWindow ?? OLLAMA_DEFAULT_CONTEXT_WINDOW,
+    maxTokens: OLLAMA_DEFAULT_MAX_TOKENS,
+  };
+}
+
+/** Fetch the model list from a running Ollama instance. */
+export async function fetchOllamaModels(
+  baseUrl: string,
+): Promise<{ reachable: boolean; models: OllamaTagModel[] }> {
+  try {
+    const apiBase = resolveOllamaApiBase(baseUrl);
+    const response = await fetch(`${apiBase}/api/tags`, {
+      signal: AbortSignal.timeout(5000),
+    });
+    if (!response.ok) {
+      return { reachable: true, models: [] };
+    }
+    const data = (await response.json()) as OllamaTagsResponse;
+    const models = (data.models ?? []).filter((m) => m.name);
+    return { reachable: true, models };
+  } catch {
+    return { reachable: false, models: [] };
+  }
+}
--- a/src/commands/auth-choice-options.test.ts
+++ b/src/commands/auth-choice-options.test.ts
@@ -42,6 +42,7 @@ describe("buildAuthChoiceOptions", () => {
      "byteplus-api-key",
      "vllm",
      "opencode-go",
+      "ollama",
    ]) {
      expect(options.some((opt) => opt.value === value)).toBe(true);
    }
@@ -93,4 +94,15 @@ describe("buildAuthChoiceOptions", () => {
    expect(openCodeGroup?.options.some((opt) => opt.value === "opencode-zen")).toBe(true);
    expect(openCodeGroup?.options.some((opt) => opt.value === "opencode-go")).toBe(true);
  });
+
+  it("shows Ollama in grouped provider selection", () => {
+    const { groups } = buildAuthChoiceGroups({
+      store: EMPTY_STORE,
+      includeSkip: false,
+    });
+    const ollamaGroup = groups.find((group) => group.value === "ollama");
+
+    expect(ollamaGroup).toBeDefined();
+    expect(ollamaGroup?.options.some((opt) => opt.value === "ollama")).toBe(true);
+  });
 });
--- a/src/commands/auth-choice-options.ts
+++ b/src/commands/auth-choice-options.ts
@@ -47,6 +47,12 @@ const AUTH_CHOICE_GROUP_DEFS: {
    hint: "Local/self-hosted OpenAI-compatible",
    choices: ["vllm"],
  },
+  {
+    value: "ollama",
+    label: "Ollama",
+    hint: "Cloud and local open models",
+    choices: ["ollama"],
+  },
  {
    value: "minimax",
    label: "MiniMax",
@@ -238,6 +244,11 @@ const BASE_AUTH_CHOICE_OPTIONS: ReadonlyArray<AuthChoiceOption> = [
    label: "vLLM (custom URL + model)",
    hint: "Local/self-hosted OpenAI-compatible server",
  },
+  {
+    value: "ollama",
+    label: "Ollama",
+    hint: "Cloud and local open models",
+  },
  ...buildProviderAuthChoiceOptions(),
  {
    value: "moonshot-api-key-cn",
--- a/src/commands/auth-choice.apply.ollama.test.ts
+++ b/src/commands/auth-choice.apply.ollama.test.ts
@@ -0,0 +1,83 @@
+import { describe, expect, it, vi } from "vitest";
+import type { ApplyAuthChoiceParams } from "./auth-choice.apply.js";
+import { applyAuthChoiceOllama } from "./auth-choice.apply.ollama.js";
+
+type PromptAndConfigureOllama = typeof import("./ollama-setup.js").promptAndConfigureOllama;
+
+const promptAndConfigureOllama = vi.hoisted(() =>
+  vi.fn<PromptAndConfigureOllama>(async ({ cfg }) => ({
+    config: cfg,
+    defaultModelId: "qwen3.5:35b",
+  })),
+);
+const ensureOllamaModelPulled = vi.hoisted(() => vi.fn(async () => {}));
+vi.mock("./ollama-setup.js", () => ({
+  promptAndConfigureOllama,
+  ensureOllamaModelPulled,
+}));
+
+function buildParams(overrides: Partial<ApplyAuthChoiceParams> = {}): ApplyAuthChoiceParams {
+  return {
+    authChoice: "ollama",
+    config: {},
+    prompter: {} as ApplyAuthChoiceParams["prompter"],
+    runtime: {} as ApplyAuthChoiceParams["runtime"],
+    setDefaultModel: false,
+    ...overrides,
+  };
+}
+
+describe("applyAuthChoiceOllama", () => {
+  it("returns agentModelOverride when setDefaultModel is false", async () => {
+    const config = { agents: { defaults: { model: { primary: "openai/gpt-4o-mini" } } } };
+    promptAndConfigureOllama.mockResolvedValueOnce({
+      config,
+      defaultModelId: "qwen2.5-coder:7b",
+    });
+
+    const result = await applyAuthChoiceOllama(
+      buildParams({
+        config,
+        setDefaultModel: false,
+      }),
+    );
+
+    expect(result).toEqual({
+      config,
+      agentModelOverride: "ollama/qwen2.5-coder:7b",
+    });
+    // Pull is deferred — the wizard model picker handles it.
+    expect(ensureOllamaModelPulled).not.toHaveBeenCalled();
+  });
+
+  it("sets global default model and preserves fallbacks when setDefaultModel is true", async () => {
+    const config = {
+      agents: {
+        defaults: {
+          model: {
+            primary: "openai/gpt-4o-mini",
+            fallbacks: ["anthropic/claude-sonnet-4-5"],
+          },
+        },
+      },
+    };
+    promptAndConfigureOllama.mockResolvedValueOnce({
+      config,
+      defaultModelId: "qwen2.5-coder:7b",
+    });
+
+    const result = await applyAuthChoiceOllama(
+      buildParams({
+        config,
+        setDefaultModel: true,
+      }),
+    );
+
+    expect(result?.agentModelOverride).toBeUndefined();
+    expect(result?.config.agents?.defaults?.model).toEqual({
+      primary: "ollama/qwen2.5-coder:7b",
+      fallbacks: ["anthropic/claude-sonnet-4-5"],
+    });
+    expect(ensureOllamaModelPulled).toHaveBeenCalledOnce();
+  });
+});
--- a/src/commands/auth-choice.apply.ollama.ts
+++ b/src/commands/auth-choice.apply.ollama.ts
@@ -0,0 +1,31 @@
+import type { ApplyAuthChoiceParams, ApplyAuthChoiceResult } from "./auth-choice.apply.js";
+import { ensureOllamaModelPulled, promptAndConfigureOllama } from "./ollama-setup.js";
+import { applyAgentDefaultModelPrimary } from "./onboard-auth.config-shared.js";
+
+export async function applyAuthChoiceOllama(
+  params: ApplyAuthChoiceParams,
+): Promise<ApplyAuthChoiceResult | null> {
+  if (params.authChoice !== "ollama") {
+    return null;
+  }
+
+  const { config, defaultModelId } = await promptAndConfigureOllama({
+    cfg: params.config,
+    prompter: params.prompter,
+    agentDir: params.agentDir,
+  });
+
+  // Set an Ollama default so the model picker pre-selects an Ollama model.
+  const defaultModel = `ollama/${defaultModelId}`;
+  const configWithDefault = applyAgentDefaultModelPrimary(config, defaultModel);
+
+  if (!params.setDefaultModel) {
+    // Defer pulling: the interactive wizard will show a model picker next,
+    // so avoid downloading a model the user may not choose.
+    return { config, agentModelOverride: defaultModel };
+  }
+
+  await ensureOllamaModelPulled({ config: configWithDefault, prompter: params.prompter });
+
+  return { config: configWithDefault };
+}
--- a/src/commands/auth-choice.apply.ts
+++ b/src/commands/auth-choice.apply.ts
@@ -9,6 +9,7 @@ import { applyAuthChoiceGitHubCopilot } from "./auth-choice.apply.github-copilot
 import { applyAuthChoiceGoogleGeminiCli } from "./auth-choice.apply.google-gemini-cli.js";
 import { applyAuthChoiceMiniMax } from "./auth-choice.apply.minimax.js";
 import { applyAuthChoiceOAuth } from "./auth-choice.apply.oauth.js";
+import { applyAuthChoiceOllama } from "./auth-choice.apply.ollama.js";
 import { applyAuthChoiceOpenAI } from "./auth-choice.apply.openai.js";
 import { applyAuthChoiceQwenPortal } from "./auth-choice.apply.qwen-portal.js";
 import { applyAuthChoiceVllm } from "./auth-choice.apply.vllm.js";
@@ -38,6 +39,7 @@ export async function applyAuthChoice(
  const handlers: Array<(p: ApplyAuthChoiceParams) => Promise<ApplyAuthChoiceResult | null>> = [
    applyAuthChoiceAnthropic,
    applyAuthChoiceVllm,
+    applyAuthChoiceOllama,
    applyAuthChoiceOpenAI,
    applyAuthChoiceOAuth,
    applyAuthChoiceApiProviders,
--- a/src/commands/auth-choice.preferred-provider.ts
+++ b/src/commands/auth-choice.preferred-provider.ts
@@ -7,6 +7,7 @@ const PREFERRED_PROVIDER_BY_AUTH_CHOICE: Partial<Record<AuthChoice, string>> = {
  token: "anthropic",
  apiKey: "anthropic",
  vllm: "vllm",
+  ollama: "ollama",
  "openai-codex": "openai-codex",
  "codex-cli": "openai-codex",
  chutes: "chutes",
--- a/src/commands/auth-choice.test.ts
+++ b/src/commands/auth-choice.test.ts
@@ -22,6 +22,7 @@ import {
 } from "./test-wizard-helpers.js";

 type DetectZaiEndpoint = typeof import("./zai-endpoint-detect.js").detectZaiEndpoint;
+type PromptAndConfigureOllama = typeof import("./ollama-setup.js").promptAndConfigureOllama;

 vi.mock("../providers/github-copilot-auth.js", () => ({
  githubCopilotLoginCommand: vi.fn(async () => {}),
@@ -44,6 +45,16 @@ vi.mock("./zai-endpoint-detect.js", () => ({
  detectZaiEndpoint,
 }));

+const promptAndConfigureOllama = vi.hoisted(() =>
+  vi.fn<PromptAndConfigureOllama>(async ({ cfg }) => ({
+    config: cfg,
+    defaultModelId: "qwen3.5:35b",
+  })),
+);
+vi.mock("./ollama-setup.js", () => ({
+  promptAndConfigureOllama,
+}));
+
 type StoredAuthProfile = {
  key?: string;
  keyRef?: { source: string; provider: string; id: string };
@@ -131,6 +142,11 @@ describe("applyAuthChoice", () => {
    detectZaiEndpoint.mockResolvedValue(null);
    loginOpenAICodexOAuth.mockReset();
    loginOpenAICodexOAuth.mockResolvedValue(null);
+    promptAndConfigureOllama.mockReset();
+    promptAndConfigureOllama.mockImplementation(async ({ cfg }) => ({
+      config: cfg,
+      defaultModelId: "qwen3.5:35b",
+    }));
    await lifecycle.cleanup();
    activeStateDir = null;
  });
@@ -1350,6 +1366,7 @@ describe("resolvePreferredProviderForAuthChoice", () => {
      { authChoice: "github-copilot" as const, expectedProvider: "github-copilot" },
      { authChoice: "qwen-portal" as const, expectedProvider: "qwen-portal" },
      { authChoice: "mistral-api-key" as const, expectedProvider: "mistral" },
+      { authChoice: "ollama" as const, expectedProvider: "ollama" },
      { authChoice: "unknown" as AuthChoice, expectedProvider: undefined },
    ] as const;
    for (const scenario of scenarios) {
--- a/src/commands/ollama-setup.test.ts
+++ b/src/commands/ollama-setup.test.ts
@@ -0,0 +1,391 @@
+import { afterEach, describe, expect, it, vi } from "vitest";
+import type { RuntimeEnv } from "../runtime.js";
+import type { WizardPrompter } from "../wizard/prompts.js";
+import {
+  configureOllamaNonInteractive,
+  ensureOllamaModelPulled,
+  promptAndConfigureOllama,
+} from "./ollama-setup.js";
+
+const upsertAuthProfileWithLock = vi.hoisted(() => vi.fn(async () => {}));
+vi.mock("../agents/auth-profiles.js", () => ({
+  upsertAuthProfileWithLock,
+}));
+
+const openUrlMock = vi.hoisted(() => vi.fn(async () => false));
+vi.mock("./onboard-helpers.js", async (importOriginal) => {
+  const original = await importOriginal<typeof import("./onboard-helpers.js")>();
+  return { ...original, openUrl: openUrlMock };
+});
+
+const isRemoteEnvironmentMock = vi.hoisted(() => vi.fn(() => false));
+vi.mock("./oauth-env.js", () => ({
+  isRemoteEnvironment: isRemoteEnvironmentMock,
+}));
+
+function jsonResponse(body: unknown, status = 200): Response {
+  return new Response(JSON.stringify(body), {
+    status,
+    headers: { "Content-Type": "application/json" },
+  });
+}
+
+describe("ollama setup", () => {
+  afterEach(() => {
+    vi.unstubAllGlobals();
+    upsertAuthProfileWithLock.mockClear();
+    openUrlMock.mockClear();
+    isRemoteEnvironmentMock.mockReset().mockReturnValue(false);
+  });
+
+  it("returns suggested default model for local mode", async () => {
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("local"),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [{ name: "llama3:8b" }] }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const result = await promptAndConfigureOllama({ cfg: {}, prompter });
+
+    expect(result.defaultModelId).toBe("glm-4.7-flash");
+  });
+
+  it("returns suggested default model for remote mode", async () => {
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("remote"),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [{ name: "llama3:8b" }] }))
+      .mockResolvedValueOnce(jsonResponse({ username: "testuser" }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const result = await promptAndConfigureOllama({ cfg: {}, prompter });
+
+    expect(result.defaultModelId).toBe("kimi-k2.5:cloud");
+  });
+
+  it("mode selection affects model ordering (local)", async () => {
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("local"),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(
+        jsonResponse({ models: [{ name: "llama3:8b" }, { name: "glm-4.7-flash" }] }),
+      );
+    vi.stubGlobal("fetch", fetchMock);
+
+    const result = await promptAndConfigureOllama({ cfg: {}, prompter });
+
+    expect(result.defaultModelId).toBe("glm-4.7-flash");
+    const modelIds = result.config.models?.providers?.ollama?.models?.map((m) => m.id);
+    expect(modelIds?.[0]).toBe("glm-4.7-flash");
+    expect(modelIds).toContain("llama3:8b");
+  });
+
+  it("cloud+local mode triggers /api/me check and opens sign-in URL", async () => {
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("remote"),
+      confirm: vi.fn().mockResolvedValueOnce(true),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [{ name: "llama3:8b" }] }))
+      .mockResolvedValueOnce(
+        jsonResponse({ error: "not signed in", signin_url: "https://ollama.com/signin" }, 401),
+      )
+      .mockResolvedValueOnce(jsonResponse({ username: "testuser" }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    await promptAndConfigureOllama({ cfg: {}, prompter });
+
+    expect(openUrlMock).toHaveBeenCalledWith("https://ollama.com/signin");
+    expect(prompter.confirm).toHaveBeenCalled();
+  });
+
+  it("cloud+local mode does not open browser in remote environment", async () => {
+    isRemoteEnvironmentMock.mockReturnValue(true);
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("remote"),
+      confirm: vi.fn().mockResolvedValueOnce(true),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [{ name: "llama3:8b" }] }))
+      .mockResolvedValueOnce(
+        jsonResponse({ error: "not signed in", signin_url: "https://ollama.com/signin" }, 401),
+      )
+      .mockResolvedValueOnce(jsonResponse({ username: "testuser" }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    await promptAndConfigureOllama({ cfg: {}, prompter });
+
+    expect(openUrlMock).not.toHaveBeenCalled();
+  });
+
+  it("local mode does not trigger cloud auth", async () => {
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("local"),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [{ name: "llama3:8b" }] }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    await promptAndConfigureOllama({ cfg: {}, prompter });
+
+    expect(fetchMock).toHaveBeenCalledTimes(1);
+    expect(fetchMock.mock.calls[0][0]).toContain("/api/tags");
+  });
+
+  it("suggested models appear first in model list (cloud+local)", async () => {
+    const prompter = {
+      text: vi.fn().mockResolvedValueOnce("http://127.0.0.1:11434"),
+      select: vi.fn().mockResolvedValueOnce("remote"),
+      note: vi.fn(async () => undefined),
+    } as unknown as WizardPrompter;
+
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(
+        jsonResponse({
+          models: [{ name: "llama3:8b" }, { name: "glm-4.7-flash" }, { name: "deepseek-r1:14b" }],
+        }),
+      )
+      .mockResolvedValueOnce(jsonResponse({ username: "testuser" }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const result = await promptAndConfigureOllama({ cfg: {}, prompter });
+    const modelIds = result.config.models?.providers?.ollama?.models?.map((m) => m.id);
+
+    expect(modelIds).toEqual([
+      "kimi-k2.5:cloud",
+      "minimax-m2.5:cloud",
+      "glm-5:cloud",
+      "llama3:8b",
+      "glm-4.7-flash",
+      "deepseek-r1:14b",
+    ]);
+  });
+
+  describe("ensureOllamaModelPulled", () => {
+    it("pulls model when not available locally", async () => {
+      const progress = { update: vi.fn(), stop: vi.fn() };
+      const prompter = {
+        progress: vi.fn(() => progress),
+      } as unknown as WizardPrompter;
+
+      const fetchMock = vi
+        .fn()
+        // /api/tags — model not present
+        .mockResolvedValueOnce(jsonResponse({ models: [{ name: "llama3:8b" }] }))
+        // /api/pull
+        .mockResolvedValueOnce(new Response('{"status":"success"}\n', { status: 200 }));
+      vi.stubGlobal("fetch", fetchMock);
+
+      await ensureOllamaModelPulled({
+        config: {
+          agents: { defaults: { model: { primary: "ollama/glm-4.7-flash" } } },
+          models: { providers: { ollama: { baseUrl: "http://127.0.0.1:11434", models: [] } } },
+        },
+        prompter,
+      });
+
+      expect(fetchMock).toHaveBeenCalledTimes(2);
+      expect(fetchMock.mock.calls[1][0]).toContain("/api/pull");
+    });
+
+    it("skips pull when model is already available", async () => {
+      const prompter = {} as unknown as WizardPrompter;
+
+      const fetchMock = vi
+        .fn()
+        .mockResolvedValueOnce(jsonResponse({ models: [{ name: "glm-4.7-flash" }] }));
+      vi.stubGlobal("fetch", fetchMock);
+
+      await ensureOllamaModelPulled({
+        config: {
+          agents: { defaults: { model: { primary: "ollama/glm-4.7-flash" } } },
+          models: { providers: { ollama: { baseUrl: "http://127.0.0.1:11434", models: [] } } },
+        },
+        prompter,
+      });
+
+      expect(fetchMock).toHaveBeenCalledTimes(1);
+    });
+
+    it("skips pull for cloud models", async () => {
+      const prompter = {} as unknown as WizardPrompter;
+      const fetchMock = vi.fn();
+      vi.stubGlobal("fetch", fetchMock);
+
+      await ensureOllamaModelPulled({
+        config: {
+          agents: { defaults: { model: { primary: "ollama/kimi-k2.5:cloud" } } },
+          models: { providers: { ollama: { baseUrl: "http://127.0.0.1:11434", models: [] } } },
+        },
+        prompter,
+      });
+
+      expect(fetchMock).not.toHaveBeenCalled();
+    });
+
+    it("skips when model is not an ollama model", async () => {
+      const prompter = {} as unknown as WizardPrompter;
+      const fetchMock = vi.fn();
+      vi.stubGlobal("fetch", fetchMock);
+
+      await ensureOllamaModelPulled({
+        config: {
+          agents: { defaults: { model: { primary: "openai/gpt-4o" } } },
+        },
+        prompter,
+      });
+
+      expect(fetchMock).not.toHaveBeenCalled();
+    });
+  });
+
+  it("uses discovered model when requested non-interactive download fails", async () => {
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [{ name: "qwen2.5-coder:7b" }] }))
+      .mockResolvedValueOnce(new Response('{"error":"disk full"}\n', { status: 200 }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const runtime = {
+      log: vi.fn(),
+      error: vi.fn(),
+      exit: vi.fn(),
+    } as unknown as RuntimeEnv;
+
+    const result = await configureOllamaNonInteractive({
+      nextConfig: {
+        agents: {
+          defaults: {
+            model: {
+              primary: "openai/gpt-4o-mini",
+              fallbacks: ["anthropic/claude-sonnet-4-5"],
+            },
+          },
+        },
+      },
+      opts: {
+        customBaseUrl: "http://127.0.0.1:11434",
+        customModelId: "missing-model",
+      },
+      runtime,
+    });
+
+    expect(runtime.error).toHaveBeenCalledWith("Download failed: disk full");
+    expect(result.agents?.defaults?.model).toEqual({
+      primary: "ollama/qwen2.5-coder:7b",
+      fallbacks: ["anthropic/claude-sonnet-4-5"],
+    });
+  });
+
+  it("normalizes ollama/ prefix in non-interactive custom model download", async () => {
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ models: [] }))
+      .mockResolvedValueOnce(new Response('{"status":"success"}\n', { status: 200 }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const runtime = {
+      log: vi.fn(),
+      error: vi.fn(),
+      exit: vi.fn(),
+    } as unknown as RuntimeEnv;
+
+    const result = await configureOllamaNonInteractive({
+      nextConfig: {},
+      opts: {
+        customBaseUrl: "http://127.0.0.1:11434",
+        customModelId: "ollama/llama3.2:latest",
+      },
+      runtime,
+    });
+
+    const pullRequest = fetchMock.mock.calls[1]?.[1];
+    expect(JSON.parse(String(pullRequest?.body))).toEqual({ name: "llama3.2:latest" });
+    expect(result.agents?.defaults?.model).toEqual(
+      expect.objectContaining({ primary: "ollama/llama3.2:latest" }),
+    );
+  });
+
+  it("accepts cloud models in non-interactive mode without pulling", async () => {
+    const fetchMock = vi.fn().mockResolvedValueOnce(jsonResponse({ models: [] }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const runtime = {
+      log: vi.fn(),
+      error: vi.fn(),
+      exit: vi.fn(),
+    } as unknown as RuntimeEnv;
+
+    const result = await configureOllamaNonInteractive({
+      nextConfig: {},
+      opts: {
+        customBaseUrl: "http://127.0.0.1:11434",
+        customModelId: "kimi-k2.5:cloud",
+      },
+      runtime,
+    });
+
+    expect(fetchMock).toHaveBeenCalledTimes(1);
+    expect(result.models?.providers?.ollama?.models?.map((model) => model.id)).toContain(
+      "kimi-k2.5:cloud",
+    );
+    expect(result.agents?.defaults?.model).toEqual(
+      expect.objectContaining({ primary: "ollama/kimi-k2.5:cloud" }),
+    );
+  });
+
+  it("exits when Ollama is unreachable", async () => {
+    const fetchMock = vi.fn().mockRejectedValueOnce(new Error("connect ECONNREFUSED"));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const runtime = {
+      log: vi.fn(),
+      error: vi.fn(),
+      exit: vi.fn(),
+    } as unknown as RuntimeEnv;
+    const nextConfig = {};
+
+    const result = await configureOllamaNonInteractive({
+      nextConfig,
+      opts: {
+        customBaseUrl: "http://127.0.0.1:11435",
+        customModelId: "llama3.2:latest",
+      },
+      runtime,
+    });
+
+    expect(runtime.error).toHaveBeenCalledWith(
+      expect.stringContaining("Ollama could not be reached at http://127.0.0.1:11435."),
+    );
+    expect(runtime.exit).toHaveBeenCalledWith(1);
+    expect(result).toBe(nextConfig);
+  });
+});
--- a/src/commands/ollama-setup.ts
+++ b/src/commands/ollama-setup.ts
@@ -0,0 +1,511 @@
+import { upsertAuthProfileWithLock } from "../agents/auth-profiles.js";
+import {
+  OLLAMA_DEFAULT_BASE_URL,
+  buildOllamaModelDefinition,
+  fetchOllamaModels,
+  resolveOllamaApiBase,
+} from "../agents/ollama-models.js";
+import type { OpenClawConfig } from "../config/config.js";
+import type { RuntimeEnv } from "../runtime.js";
+import { WizardCancelledError, type WizardPrompter } from "../wizard/prompts.js";
+import { isRemoteEnvironment } from "./oauth-env.js";
+import { applyAgentDefaultModelPrimary } from "./onboard-auth.config-shared.js";
+import { openUrl } from "./onboard-helpers.js";
+import type { OnboardMode, OnboardOptions } from "./onboard-types.js";
+
+export { OLLAMA_DEFAULT_BASE_URL } from "../agents/ollama-models.js";
+export const OLLAMA_DEFAULT_MODEL = "glm-4.7-flash";
+
+const OLLAMA_SUGGESTED_MODELS_LOCAL = ["glm-4.7-flash"];
+const OLLAMA_SUGGESTED_MODELS_CLOUD = [
+  "kimi-k2.5:cloud",
+  "minimax-m2.5:cloud",
+  "glm-5:cloud",
+];
+
+function normalizeOllamaModelName(value: string | undefined): string | undefined {
+  const trimmed = value?.trim();
+  if (!trimmed) {
+    return undefined;
+  }
+  if (trimmed.toLowerCase().startsWith("ollama/")) {
+    const withoutPrefix = trimmed.slice("ollama/".length).trim();
+    return withoutPrefix || undefined;
+  }
+  return trimmed;
+}
+
+function isOllamaCloudModel(modelName: string | undefined): boolean {
+  return Boolean(modelName?.trim().toLowerCase().endsWith(":cloud"));
+}
+
+function formatOllamaPullStatus(status: string): { text: string; hidePercent: boolean } {
+  const trimmed = status.trim();
+  const partStatusMatch = trimmed.match(/^([a-z-]+)\s+(?:sha256:)?[a-f0-9]{8,}$/i);
+  if (partStatusMatch) {
+    return { text: `${partStatusMatch[1]} part`, hidePercent: false };
+  }
+  if (/^verifying\b.*\bdigest\b/i.test(trimmed)) {
+    return { text: "verifying digest", hidePercent: true };
+  }
+  return { text: trimmed, hidePercent: false };
+}
+
+type OllamaCloudAuthResult = {
+  signedIn: boolean;
+  signinUrl?: string;
+};
+
+/** Check if the user is signed in to Ollama cloud via /api/me. */
+async function checkOllamaCloudAuth(baseUrl: string): Promise<OllamaCloudAuthResult> {
+  try {
+    const apiBase = resolveOllamaApiBase(baseUrl);
+    const response = await fetch(`${apiBase}/api/me`, {
+      method: "POST",
+      signal: AbortSignal.timeout(5000),
+    });
+    if (response.status === 401) {
+      // 401 body contains { error, signin_url }
+      const data = (await response.json()) as { signin_url?: string };
+      return { signedIn: false, signinUrl: data.signin_url };
+    }
+    if (!response.ok) {
+      return { signedIn: false };
+    }
+    return { signedIn: true };
+  } catch {
+    // /api/me not supported or unreachable — fail closed so cloud mode
+    // doesn't silently skip auth; the caller handles the fallback.
+    return { signedIn: false };
+  }
+}
+
+type OllamaPullChunk = {
+  status?: string;
+  total?: number;
+  completed?: number;
+  error?: string;
+};
+
+type OllamaPullFailureKind = "http" | "no-body" | "chunk-error" | "network";
+type OllamaPullResult =
+  | { ok: true }
+  | {
+      ok: false;
+      kind: OllamaPullFailureKind;
+      message: string;
+    };
+
+async function pullOllamaModelCore(params: {
+  baseUrl: string;
+  modelName: string;
+  onStatus?: (status: string, percent: number | null) => void;
+}): Promise<OllamaPullResult> {
+  const { onStatus } = params;
+  const baseUrl = resolveOllamaApiBase(params.baseUrl);
+  const modelName = normalizeOllamaModelName(params.modelName) ?? params.modelName.trim();
+  try {
+    const response = await fetch(`${baseUrl}/api/pull`, {
+      method: "POST",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({ name: modelName }),
+    });
+    if (!response.ok) {
+      return {
+        ok: false,
+        kind: "http",
+        message: `Failed to download ${modelName} (HTTP ${response.status})`,
+      };
+    }
+    if (!response.body) {
+      return {
+        ok: false,
+        kind: "no-body",
+        message: `Failed to download ${modelName} (no response body)`,
+      };
+    }
+
+    const reader = response.body.getReader();
+    const decoder = new TextDecoder();
+    let buffer = "";
+    const layers = new Map<string, { total: number; completed: number }>();
+
+    const parseLine = (line: string): OllamaPullResult => {
+      const trimmed = line.trim();
+      if (!trimmed) {
+        return { ok: true };
+      }
+      try {
+        const chunk = JSON.parse(trimmed) as OllamaPullChunk;
+        if (chunk.error) {
+          return {
+            ok: false,
+            kind: "chunk-error",
+            message: `Download failed: ${chunk.error}`,
+          };
+        }
+        if (!chunk.status) {
+          return { ok: true };
+        }
+        if (chunk.total && chunk.completed !== undefined) {
+          layers.set(chunk.status, { total: chunk.total, completed: chunk.completed });
+          let totalSum = 0;
+          let completedSum = 0;
+          for (const layer of layers.values()) {
+            totalSum += layer.total;
+            completedSum += layer.completed;
+          }
+          const percent = totalSum > 0 ? Math.round((completedSum / totalSum) * 100) : null;
+          onStatus?.(chunk.status, percent);
+        } else {
+          onStatus?.(chunk.status, null);
+        }
+      } catch {
+        // Ignore malformed lines from streaming output.
+      }
+      return { ok: true };
+    };
+
+    for (;;) {
+      const { done, value } = await reader.read();
+      if (done) {
+        break;
+      }
+      buffer += decoder.decode(value, { stream: true });
+      const lines = buffer.split("\n");
+      buffer = lines.pop() ?? "";
+      for (const line of lines) {
+        const parsed = parseLine(line);
+        if (!parsed.ok) {
+          return parsed;
+        }
+      }
+    }
+
+    const trailing = buffer.trim();
+    if (trailing) {
+      const parsed = parseLine(trailing);
+      if (!parsed.ok) {
+        return parsed;
+      }
+    }
+
+    return { ok: true };
+  } catch (err) {
+    const reason = err instanceof Error ? err.message : String(err);
+    return {
+      ok: false,
+      kind: "network",
+      message: `Failed to download ${modelName}: ${reason}`,
+    };
+  }
+}
+
+/** Pull a model from Ollama, streaming progress updates. */
+async function pullOllamaModel(
+  baseUrl: string,
+  modelName: string,
+  prompter: WizardPrompter,
+): Promise<boolean> {
+  const spinner = prompter.progress(`Downloading ${modelName}...`);
+  const result = await pullOllamaModelCore({
+    baseUrl,
+    modelName,
+    onStatus: (status, percent) => {
+      const displayStatus = formatOllamaPullStatus(status);
+      if (displayStatus.hidePercent) {
+        spinner.update(`Downloading ${modelName} - ${displayStatus.text}`);
+      } else {
+        spinner.update(`Downloading ${modelName} - ${displayStatus.text} - ${percent ?? 0}%`);
+      }
+    },
+  });
+  if (!result.ok) {
+    spinner.stop(result.message);
+    return false;
+  }
+  spinner.stop(`Downloaded ${modelName}`);
+  return true;
+}
+
+async function pullOllamaModelNonInteractive(
+  baseUrl: string,
+  modelName: string,
+  runtime: RuntimeEnv,
+): Promise<boolean> {
+  runtime.log(`Downloading ${modelName}...`);
+  const result = await pullOllamaModelCore({ baseUrl, modelName });
+  if (!result.ok) {
+    runtime.error(result.message);
+    return false;
+  }
+  runtime.log(`Downloaded ${modelName}`);
+  return true;
+}
+
+function buildOllamaModelsConfig(modelNames: string[]) {
+  return modelNames.map((name) => buildOllamaModelDefinition(name));
+}
+
+function applyOllamaProviderConfig(
+  cfg: OpenClawConfig,
+  baseUrl: string,
+  modelNames: string[],
+): OpenClawConfig {
+  return {
+    ...cfg,
+    models: {
+      ...cfg.models,
+      mode: cfg.models?.mode ?? "merge",
+      providers: {
+        ...cfg.models?.providers,
+        ollama: {
+          baseUrl,
+          api: "ollama",
+          apiKey: "OLLAMA_API_KEY", // pragma: allowlist secret
+          models: buildOllamaModelsConfig(modelNames),
+        },
+      },
+    },
+  };
+}
+
+async function storeOllamaCredential(agentDir?: string): Promise<void> {
+  await upsertAuthProfileWithLock({
+    profileId: "ollama:default",
+    credential: { type: "api_key", provider: "ollama", key: "ollama-local" },
+    agentDir,
+  });
+}
+
+/**
+ * Interactive: prompt for base URL, discover models, configure provider.
+ * Model selection is handled by the standard model picker downstream.
+ */
+export async function promptAndConfigureOllama(params: {
+  cfg: OpenClawConfig;
+  prompter: WizardPrompter;
+  agentDir?: string;
+}): Promise<{ config: OpenClawConfig; defaultModelId: string }> {
+  const { prompter } = params;
+
+  // 1. Prompt base URL
+  const baseUrlRaw = await prompter.text({
+    message: "Ollama base URL",
+    initialValue: OLLAMA_DEFAULT_BASE_URL,
+    placeholder: OLLAMA_DEFAULT_BASE_URL,
+    validate: (value) => (value?.trim() ? undefined : "Required"),
+  });
+  const configuredBaseUrl = String(baseUrlRaw ?? "")
+    .trim()
+    .replace(/\/+$/, "");
+  const baseUrl = resolveOllamaApiBase(configuredBaseUrl);
+
+  // 2. Check reachability
+  const { reachable, models } = await fetchOllamaModels(baseUrl);
+  const modelNames = models.map((m) => m.name);
+
+  if (!reachable) {
+    await prompter.note(
+      [
+        `Ollama could not be reached at ${baseUrl}.`,
+        "Download it at https://ollama.com/download",
+        "",
+        "Start Ollama and re-run onboarding.",
+      ].join("\n"),
+      "Ollama",
+    );
+    throw new WizardCancelledError("Ollama not reachable");
+  }
+
+  // 3. Mode selection
+  const mode = (await prompter.select({
+    message: "Ollama mode",
+    options: [
+      { value: "remote", label: "Cloud + Local", hint: "Ollama cloud models + local models" },
+      { value: "local", label: "Local", hint: "Local models only" },
+    ],
+  })) as OnboardMode;
+
+  // 4. Cloud auth — check /api/me upfront for remote (cloud+local) mode
+  let cloudAuthVerified = false;
+  if (mode === "remote") {
+    const authResult = await checkOllamaCloudAuth(baseUrl);
+    if (!authResult.signedIn) {
+      if (authResult.signinUrl) {
+        if (!isRemoteEnvironment()) {
+          await openUrl(authResult.signinUrl);
+        }
+        await prompter.note(
+          ["Sign in to Ollama Cloud:", authResult.signinUrl].join("\n"),
+          "Ollama Cloud",
+        );
+        const confirmed = await prompter.confirm({
+          message: "Have you signed in?",
+        });
+        if (!confirmed) {
+          throw new WizardCancelledError("Ollama cloud sign-in cancelled");
+        }
+        // Re-check after user claims sign-in
+        const recheck = await checkOllamaCloudAuth(baseUrl);
+        if (!recheck.signedIn) {
+          throw new WizardCancelledError("Ollama cloud sign-in required");
+        }
+        cloudAuthVerified = true;
+      } else {
+        // No signin URL available (older server, unreachable /api/me, or custom gateway).
+        await prompter.note(
+          [
+            "Could not verify Ollama Cloud authentication.",
+            "Cloud models may not work until you sign in at https://ollama.com.",
+          ].join("\n"),
+          "Ollama Cloud",
+        );
+        const continueAnyway = await prompter.confirm({
+          message: "Continue without cloud auth?",
+        });
+        if (!continueAnyway) {
+          throw new WizardCancelledError("Ollama cloud auth could not be verified");
+        }
+        // Cloud auth unverified — fall back to local defaults so the model
+        // picker doesn't steer toward cloud models that may fail.
+      }
+    } else {
+      cloudAuthVerified = true;
+    }
+  }
+
+  // 5. Model ordering — suggested models first.
+  // Use cloud defaults only when auth was actually verified; otherwise fall
+  // back to local defaults so the user isn't steered toward cloud models
+  // that may fail at runtime.
+  const suggestedModels =
+    mode === "local" || !cloudAuthVerified
+      ? OLLAMA_SUGGESTED_MODELS_LOCAL
+      : OLLAMA_SUGGESTED_MODELS_CLOUD;
+  const orderedModelNames = [
+    ...suggestedModels,
+    ...modelNames.filter((name) => !suggestedModels.includes(name)),
+  ];
+
+  await storeOllamaCredential(params.agentDir);
+
+  const defaultModelId = suggestedModels[0] ?? OLLAMA_DEFAULT_MODEL;
+  const config = applyOllamaProviderConfig(params.cfg, baseUrl, orderedModelNames);
+  return { config, defaultModelId };
+}
+
+/** Non-interactive: auto-discover models and configure provider. */
+export async function configureOllamaNonInteractive(params: {
+  nextConfig: OpenClawConfig;
+  opts: OnboardOptions;
+  runtime: RuntimeEnv;
+}): Promise<OpenClawConfig> {
+  const { opts, runtime } = params;
+  const configuredBaseUrl = (opts.customBaseUrl?.trim() || OLLAMA_DEFAULT_BASE_URL).replace(
+    /\/+$/,
+    "",
+  );
+  const baseUrl = resolveOllamaApiBase(configuredBaseUrl);
+
+  const { reachable, models } = await fetchOllamaModels(baseUrl);
+  const modelNames = models.map((m) => m.name);
+  const explicitModel = normalizeOllamaModelName(opts.customModelId);
+
+  if (!reachable) {
+    runtime.error(
+      [
+        `Ollama could not be reached at ${baseUrl}.`,
+        "Download it at https://ollama.com/download",
+      ].join("\n"),
+    );
+    runtime.exit(1);
+    return params.nextConfig;
+  }
+
+  await storeOllamaCredential();
+
+  // Apply local suggested model ordering.
+  const suggestedModels = OLLAMA_SUGGESTED_MODELS_LOCAL;
+  const orderedModelNames = [
+    ...suggestedModels,
+    ...modelNames.filter((name) => !suggestedModels.includes(name)),
+  ];
+
+  const requestedDefaultModelId = explicitModel ?? suggestedModels[0];
+  let pulledRequestedModel = false;
+  const availableModelNames = new Set(modelNames);
+  const requestedCloudModel = isOllamaCloudModel(requestedDefaultModelId);
+
+  if (requestedCloudModel) {
+    availableModelNames.add(requestedDefaultModelId);
+  }
+
+  // Pull if model not in discovered list and Ollama is reachable
+  if (!requestedCloudModel && !modelNames.includes(requestedDefaultModelId)) {
+    pulledRequestedModel = await pullOllamaModelNonInteractive(
+      baseUrl,
+      requestedDefaultModelId,
+      runtime,
+    );
+    if (pulledRequestedModel) {
+      availableModelNames.add(requestedDefaultModelId);
+    }
+  }
+
+  let allModelNames = orderedModelNames;
+  let defaultModelId = requestedDefaultModelId;
+  if ((pulledRequestedModel || requestedCloudModel) && !allModelNames.includes(requestedDefaultModelId)) {
+    allModelNames = [...allModelNames, requestedDefaultModelId];
+  }
+  if (!availableModelNames.has(requestedDefaultModelId)) {
+    if (availableModelNames.size > 0) {
+      const firstAvailableModel =
+        allModelNames.find((name) => availableModelNames.has(name)) ??
+        Array.from(availableModelNames)[0];
+      defaultModelId = firstAvailableModel;
+      runtime.log(
+        `Ollama model ${requestedDefaultModelId} was not available; using ${defaultModelId} instead.`,
+      );
+    } else {
+      runtime.error(
+        [
+          `No Ollama models are available at ${baseUrl}.`,
+          "Pull a model first, then re-run onboarding.",
+        ].join("\n"),
+      );
+      runtime.exit(1);
+      return params.nextConfig;
+    }
+  }
+
+  const config = applyOllamaProviderConfig(params.nextConfig, baseUrl, allModelNames);
+  const modelRef = `ollama/${defaultModelId}`;
+  runtime.log(`Default Ollama model: ${defaultModelId}`);
+  return applyAgentDefaultModelPrimary(config, modelRef);
+}
+
+/** Pull the configured default Ollama model if it isn't already available locally. */
+export async function ensureOllamaModelPulled(params: {
+  config: OpenClawConfig;
+  prompter: WizardPrompter;
+}): Promise<void> {
+  const modelCfg = params.config.agents?.defaults?.model;
+  const modelId = typeof modelCfg === "string" ? modelCfg : modelCfg?.primary;
+  if (!modelId?.startsWith("ollama/")) {
+    return;
+  }
+  const baseUrl = params.config.models?.providers?.ollama?.baseUrl ?? OLLAMA_DEFAULT_BASE_URL;
+  const modelName = modelId.slice("ollama/".length);
+  if (isOllamaCloudModel(modelName)) {
+    return;
+  }
+  const { models } = await fetchOllamaModels(baseUrl);
+  if (models.some((m) => m.name === modelName)) {
+    return;
+  }
+  const pulled = await pullOllamaModel(baseUrl, modelName, params.prompter);
+  if (!pulled) {
+    throw new WizardCancelledError("Failed to download selected Ollama model");
+  }
+}
--- a/src/commands/onboard-non-interactive/local/auth-choice.ts
+++ b/src/commands/onboard-non-interactive/local/auth-choice.ts
@@ -10,6 +10,7 @@ import { normalizeSecretInputModeInput } from "../../auth-choice.apply-helpers.j
 import { buildTokenProfileId, validateAnthropicSetupToken } from "../../auth-token.js";
 import { applyGoogleGeminiModelDefault } from "../../google-gemini-model-default.js";
 import { applyPrimaryModel } from "../../model-picker.js";
+import { configureOllamaNonInteractive } from "../../ollama-setup.js";
 import {
  applyAuthProfileConfig,
  applyCloudflareAiGatewayConfig,
@@ -174,6 +175,10 @@ export async function applyNonInteractiveAuthChoice(params: {
    return null;
  }

+  if (authChoice === "ollama") {
+    return configureOllamaNonInteractive({ nextConfig, opts, runtime });
+  }
+
  if (authChoice === "apiKey") {
    const resolved = await resolveApiKey({
      provider: "anthropic",
--- a/src/commands/onboard-types.ts
+++ b/src/commands/onboard-types.ts
@@ -10,6 +10,7 @@ export type AuthChoice =
  | "token"
  | "chutes"
  | "vllm"
+  | "ollama"
  | "openai-codex"
  | "openai-api-key"
  | "openrouter-api-key"
@@ -59,6 +60,7 @@ export type AuthChoiceGroupId =
  | "anthropic"
  | "chutes"
  | "vllm"
+  | "ollama"
  | "google"
  | "copilot"
  | "openrouter"
--- a/src/memory/embeddings-ollama.ts
+++ b/src/memory/embeddings-ollama.ts
@@ -1,4 +1,5 @@
 import { resolveEnvApiKey } from "../agents/model-auth.js";
+import { resolveOllamaApiBase } from "../agents/ollama-models.js";
 import { formatErrorMessage } from "../infra/errors.js";
 import type { SsrFPolicy } from "../infra/net/ssrf.js";
 import { normalizeOptionalSecretInput } from "../utils/normalize-secret-input.js";
@@ -17,7 +18,6 @@ export type OllamaEmbeddingClient = {
 type OllamaEmbeddingClientConfig = Omit<OllamaEmbeddingClient, "embedBatch">;

 export const DEFAULT_OLLAMA_EMBEDDING_MODEL = "nomic-embed-text";
-const DEFAULT_OLLAMA_BASE_URL = "http://127.0.0.1:11434";

 function sanitizeAndNormalizeEmbedding(vec: number[]): number[] {
  const sanitized = vec.map((value) => (Number.isFinite(value) ? value : 0));
@@ -36,14 +36,6 @@ function normalizeOllamaModel(model: string): string {
  });
 }

-function resolveOllamaApiBase(configuredBaseUrl?: string): string {
-  if (!configuredBaseUrl) {
-    return DEFAULT_OLLAMA_BASE_URL;
-  }
-  const trimmed = configuredBaseUrl.replace(/\/+$/, "");
-  return trimmed.replace(/\/v1$/i, "");
-}
-
 function resolveOllamaApiKey(options: EmbeddingProviderOptions): string | undefined {
  const remoteApiKey = resolveMemorySecretInputString({
    value: options.remote?.apiKey,
--- a/src/wizard/onboarding.ts
+++ b/src/wizard/onboarding.ts
@@ -442,13 +442,17 @@ export async function runOnboardingWizard(
      config: nextConfig,
      prompter,
      runtime,
-      setDefaultModel: true,
+      setDefaultModel: !(authChoiceFromPrompt && authChoice === "ollama"),
      opts: {
        tokenProvider: opts.tokenProvider,
        token: opts.authChoice === "apiKey" && opts.token ? opts.token : undefined,
      },
    });
    nextConfig = authResult.config;
+
+    if (authResult.agentModelOverride) {
+      nextConfig = applyPrimaryModel(nextConfig, authResult.agentModelOverride);
+    }
  }

  if (authChoiceFromPrompt && authChoice !== "custom-api-key") {
@@ -468,6 +472,11 @@ export async function runOnboardingWizard(
    }
  }

+  if (authChoice === "ollama") {
+    const { ensureOllamaModelPulled } = await import("../commands/ollama-setup.js");
+    await ensureOllamaModelPulled({ config: nextConfig, prompter });
+  }
+
  await warnIfModelConfigLooksOff(nextConfig, prompter);

  const { configureGatewayForOnboarding } = await import("./onboarding.gateway-config.js");