pi-multikey 1.9.0 → 1.11.0

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "pi-multikey",
-	"version": "1.9.0",
+	"version": "1.11.0",
 	"description": "One pi provider backed by many API keys: automatic 429 rotation, per-request key leases for concurrent subagents, and a /multikey management TUI",
 	"keywords": [
 		"pi-package",

package/presets.ts

@@ -188,44 +188,100 @@ export const PRESETS: Preset[] = [
 		id: "cline-free",
 		name: "Cline Free",
 		description:
-			"api.cline.bot — Cline account free tier: DeepSeek V4 Flash, Longcat 2.0, Laguna S 2.1, GLM 5.2 (daily per-model quota, lineup rotates)",
+			"api.cline.bot — Cline account free tier: DeepSeek V4.1 Flash, Laguna S 2.1, GLM 5.3, Solar Pro 4, Muse Spark 1.3, Union Alpha (daily per-model quota, lineup rotates)",
 		defaultPoolId: "cline",
 		baseUrl: "https://api.cline.bot/api/v1",
 		api: "openai-completions",
 		keyHint: "Cline account — use 'Sign in with Cline (device flow)', or paste an access token from ~/.cline/data/secrets.json",
+		// FREE-TIER MODEL IDS ARE EXACT — copy them from the `free` bucket of
+		// GET /api/v1/ai/cline/recommended-models, NOT from GET /models (the
+		// OpenRouter catalog dump). Entries prefixed `cline-free/` are only free
+		// under that exact id: their raw vendor id (e.g. deepseek/deepseek-v4.1-flash)
+		// routes to usage-based billing and answers 402 insufficient_credits on a
+		// zero-balance account, while the cline-free/ id rides the daily per-model
+		// quota ($0). The cline-free/ namespace is additionally gated on the
+		// X-CLIENT-TYPE header — already sent on every request via
+		// endpointHeaders() in config.ts, so no extra config is needed.
+		// Verified live 2026-09-17 (probe account, multikey's exact header set):
+		// raw deepseek-v4.1-flash → 402; cline-free/deepseek-v4.1-flash → 200, $0.
 		models: [
 			{
-				id: "deepseek/deepseek-v4-flash",
-				name: "DeepSeek V4 Flash",
+				// Feed id: cline-free/deepseek-v4.1-flash. The raw id
+				// (deepseek/deepseek-v4.1-flash) is usage-billed — see the comment above.
+				// Catalog (openrouter): ctx 1048576, out 384000 (both verified live), text+image.
+				// Effort tiers verified live via reasoning:{effort} (incl. "none" = reasoning off).
+				id: "cline-free/deepseek-v4.1-flash",
+				name: "DeepSeek V4.1 Flash (Free)",
+				reasoning: true,
+				input: ["text", "image"],
+				contextWindow: 1_048_576,
+				maxTokens: 384_000,
+				compat: { thinkingFormat: "openrouter" },
+				thinkingLevelMap: levels({ off: "none", low: "low", medium: "medium", high: "high", xhigh: "xhigh", max: "max" }),
+			},
+			{
+				// Free at its raw id (no cline-free/ prefix in the feed); 262K window measured
+				// live by the preset author. Not in any catalog (stealth) — sizes unverified.
+				id: "poolside/laguna-s-2.1:free",
+				name: "Laguna S 2.1 (Free)",
 				reasoning: true,
 				input: ["text"],
-				contextWindow: 1_000_000,
-				maxTokens: 131_072,
+				contextWindow: 262_144,
+				maxTokens: 32_768,
+			},
+			{
+				// z-ai/glm-5.2:free retired from the Cline lineup (id answers "model not found");
+				// succeeded by glm-5.3-flash. Always-on reasoning, no effort tiers; vision input.
+				id: "z-ai/glm-5.3-flash",
+				name: "GLM 5.3 (Free)",
+				reasoning: true,
+				input: ["text", "image"],
+				contextWindow: 1_048_576,
+				maxTokens: 128_000,
 			},
 			{
-				id: "meituan/longcat-2.0",
-				name: "Longcat 2.0",
+				// Feed id: cline-free/solar-pro4 (raw upstage/solar-pro4 is usage-billed).
+				// Catalog: ctx 524288 (1M rejected live: "maximum context length is 524288"),
+				// out 131072 (accepted live). Effort: none/high only (verified live).
+				id: "cline-free/solar-pro4",
+				name: "Upstage Solar Pro 4 (Free)",
 				reasoning: true,
 				input: ["text"],
-				contextWindow: 1_000_000,
+				cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
+				contextWindow: 524_288,
 				maxTokens: 131_072,
+				compat: { thinkingFormat: "openrouter" },
+				thinkingLevelMap: levels({ off: "none", high: "high" }),
 			},
 			{
-				id: "poolside/laguna-s-2.1:free",
-				name: "Laguna S 2.1 (Free)",
+				// Same Meta model as OpenCode Zen's muse-spark-1.3-contributor-free, served
+				// through Cline's chat-completions endpoint (no Responses API / session
+				// affinity needed). Always-on reasoning; text + image input.
+				// Feed id: cline-free/muse-spark-1.3-contributor (raw meta/… is usage-billed).
+				// Catalog: ctx 1048576, out 943718, effort minimal..max (no toggle = always on).
+				// NOTE: the cline-free/ id answered "not available in your region" from the
+				// PH (2026-09-17 probe); the model may be region-gated on the free tier, so
+				// catalog values are unverified against the live free route.
+				id: "cline-free/muse-spark-1.3-contributor",
+				name: "Meta Muse Spark 1.3 Contributor (Free)",
 				reasoning: true,
-				input: ["text"],
-				// Poolside doesn't publish the window; safe default, tune if needed.
-				contextWindow: 128_000,
-				maxTokens: 16_384,
+				input: ["text", "image"],
+				cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
+				contextWindow: 1_048_576,
+				maxTokens: 943_718,
+				compat: { thinkingFormat: "openrouter" },
+				thinkingLevelMap: levels({ minimal: "minimal", low: "low", medium: "medium", high: "high", xhigh: "xhigh", max: "max" }),
 			},
 			{
-				id: "z-ai/glm-5.2:free",
-				name: "GLM 5.2 (Free)",
+				// Stealth promo model; free at its raw id (no cline-free/ prefix in the feed).
+				// Not in any catalog (stealth) — sizes/reasoning copied from the live pool,
+				// unverified.
+				id: "stealth/union-alpha",
+				name: "Union Alpha",
 				reasoning: true,
 				input: ["text"],
-				contextWindow: 200_000,
-				maxTokens: 131_072,
+				contextWindow: 262_000,
+				maxTokens: 16_384,
 			},
 		],
 	},

package/tui.ts

@@ -267,7 +267,9 @@ export async function pickMany(
 				if (v.length === 0) {
 					add(theme.fg("warning", " No matches."));
 				} else {
-					for (let i = 0; i < Math.min(v.length, maxVisible); i++) {
+					const start = Math.max(0, Math.min(cursor - Math.floor(maxVisible / 2), Math.max(0, v.length - maxVisible)));
+					const end = Math.min(v.length, start + maxVisible);
+					for (let i = start; i < end; i++) {
 						const item = v[i]!;
 						const active = i === cursor;
 						const check = selected.has(item.value) ? theme.fg("accent", "✓") : " ";