free-coding-models 0.5.26 → 0.5.28

package/README.md

@@ -6,7 +6,7 @@
 
 <p align="center">
   <strong>Find the fastest free coding model in seconds</strong><br>
-  Track ~146 models across 15 trusted free or free-limited AI providers in real time<br><br>
+  Track ~191 models across 20 trusted free or free-limited AI providers in real time<br><br>
   <strong>Install Free API endpoints to your favorite AI coding tools:</strong><br>
   OpenCode CLI / Desktop / WebUI, OpenClaw, Crush, Goose, Aider, Kilo CLI, Qwen Code, OpenHands, Amp, Hermes, Continue, Cline, Xcode, Pi and more...<br><br>
   <strong>Use Kimi K2, DeepSeek V3, GPT-OSS, Qwen3, MiniMax M3, GLM, Llama 4, Gemma 4, Devstral and more — for free</strong>
@@ -16,8 +16,8 @@
   <img src="https://img.shields.io/npm/v/free-coding-models?color=3d6b00&label=npm&logo=npm" alt="npm version" width="200"><br>
   <img src="https://img.shields.io/node/v/free-coding-models?color=3d6b00&logo=node.js" alt="node version" width="200"><br>
   <img src="https://img.shields.io/npm/l/free-coding-models?color=3d6b00" alt="license" width="200"><br>
-  <img src="https://img.shields.io/badge/models-146-3d6b00?logo=nvidia" alt="models count" width="200"><br>
-  <img src="https://img.shields.io/badge/providers-15-1a56db" alt="providers count" width="200">
+  <img src="https://img.shields.io/badge/models-191-3d6b00?logo=nvidia" alt="models count" width="200"><br>
+  <img src="https://img.shields.io/badge/providers-20-1a56db" alt="providers count" width="200">
 </p>
 
 ```bash
@@ -59,7 +59,7 @@ free-coding-models
 
 ## 💡 Why this tool?
 
-There are **~153 cataloged free or free-limited coding models** across 16 vetted providers. Which one is fastest right now? Which one is actually stable versus just lucky on the last ping?
+There are **~191 cataloged free or free-limited coding models** across 20 vetted providers. Which one is fastest right now? Which one is actually stable versus just lucky on the last ping?
 
 This CLI pings them all in parallel, shows live latency, and calculates a **live Stability Score (0-100)**. Average latency alone is misleading if a model randomly spikes to 6 seconds; the stability score measures true reliability by combining **p95 latency** (30%), **jitter/variance** (30%), **spike rate** (20%), and **uptime** (20%). 
 
@@ -71,9 +71,9 @@ It then writes the model you pick directly into your coding tool's config — so
 
 ### 🟢 List of Free AI Providers
 
-Create a free account on one provider below to get started:
+Create a free account on one provider below to get started. A few providers (`Kilo`, `LLM7`, OVHcloud sandbox) can also answer without a key, with tighter shared limits.
 
-**~153 coding models** across 16 active providers, ranked by practical free-tier usefulness.
+**~191 coding models** across 20 active providers, ranked by practical free-tier usefulness.
 
 | # | Provider | Models | Tier range | Free tier | Env var |
 |---|----------|--------|-----------|-----------|--------|
@@ -83,7 +83,7 @@ Create a free account on one provider below to get started:
 | 4 | [Google AI Studio](https://aistudio.google.com/apikey) | 7 | S+ → A | Gemini free quotas vary by model/region | `GOOGLE_API_KEY` |
 | 5 | [GitHub Models](https://models.github.ai) | 15 | S+ → C | Quota depends on GitHub/Copilot tier | `GITHUB_TOKEN` |
 | 6 | [Mistral La Plateforme](https://console.mistral.ai/api-keys) | 5 | S+ → A | Experiment plan, free evaluation tier | `MISTRAL_API_KEY` |
-| 7 | [Cloudflare Workers AI](https://dash.cloudflare.com) | 13 | S+ → B | 10K neurons/day, 300 RPM (no credit card) | `CLOUDFLARE_API_TOKEN` + `CLOUDFLARE_ACCOUNT_ID` |
+| 7 | [Cloudflare Workers AI](https://dash.cloudflare.com) | 16 | S+ → B | 10K neurons/day, 300 RPM (no credit card) | `CLOUDFLARE_API_TOKEN` + `CLOUDFLARE_ACCOUNT_ID` |
 | 8 | [OpenRouter](https://openrouter.ai/keys) | 24 | S+ → C | 50 req/day free, 1K/day with $10 spend | `OPENROUTER_API_KEY` |
 | 9 | [SambaNova](https://cloud.sambanova.ai/apis) | 7 | S+ → B+ | Small developer quota, useful for light usage | `SAMBANOVA_API_KEY` |
 | 10 | [OVHcloud AI Endpoints](https://endpoints.ai.cloud.ovh.net) | 10 | S → B | 2 req/min/IP free, 400 RPM with key | `OVH_AI_ENDPOINTS_ACCESS_TOKEN` |
@@ -91,7 +91,12 @@ Create a free account on one provider below to get started:
 | 12 | [ZAI](https://z.ai) | 2 | S | Free Flash models only | `ZAI_API_KEY` |
 | 13 | [Scaleway](https://console.scaleway.com/iam/api-keys) | 10 | S+ → B | 1M free tokens | `SCALEWAY_API_KEY` |
 | 14 | [Alibaba DashScope](https://modelstudio.console.alibabacloud.com) | 11 | S+ → A+ | 1M free tokens/model, Singapore, 90 days | `DASHSCOPE_API_KEY` |
-| 15 | [OpenCode Zen](https://opencode.ai/zen) | 6 | S+ → A | Free with OpenCode account | Zen models ✨ |
+| 15 | [OpenCode Zen](https://opencode.ai/zen) | 5 | S+ → A | Free with OpenCode account | Zen models ✨ |
+| 16 | [Kilo](https://kilo.ai) | 1 | A+ | Free auto-router works without a key | optional `KILO_API_KEY` |
+| 17 | [LLM7](https://llm7.io) | 4 | S+ → B+ | Shared free tier, optional free token | optional `LLM7_API_KEY` |
+| 18 | [Routeway](https://routeway.ai) | 15 | S+ → C | Explicit `:free` zero-price models | `ROUTEWAY_API_KEY` |
+| 19 | [Novita AI](https://novita.ai) | 4 | S+ → S | Only zero-price live chat models included | `NOVITA_API_KEY` |
+| 20 | [Ollama Cloud](https://ollama.com/pricing) | 17 | S+ → A | Free cloud usage with session/weekly limits | `OLLAMA_API_KEY` |
 
 > 💡 One key is enough. Add more at any time with **`P`** inside the TUI.
 
@@ -513,7 +518,7 @@ When a tool mode is active (via `Z`), models incompatible with that tool are hig
 
 ## ✨ Features
 
-- **Parallel pings** — all ~150 API/Zen-callable models tested simultaneously via native `fetch` (~152 total cataloged models)
+- **Parallel pings** — all ~191 API/Zen-callable models tested simultaneously via native `fetch`
 - **AI benchmark columns** — `Ctrl+A` benchmarks the selected model, `Ctrl+U` benchmarks visible models, and results split cleanly into **AI Latency** plus **TPS**. Settings includes an opt-in **Startup AI Speed Scan** toggle to run the global benchmark automatically after launch.
 - **Tiny verdict indicator** — the first `❔` column mirrors the full Verdict as a compact emoji (`🟩`, `🟢`, `🟡`, `🟠`, etc.) and sorts by the same verdict order.
 - **Adaptive monitoring** — 2s burst for 60s → 10s normal → 30s idle

package/changelog/v0.5.27.md

@@ -0,0 +1,17 @@
+# Changelog v0.5.27 - 2026-06-11
+
+### Added
+- Added five newly audited free/free-limited providers discovered from `apmantza/pi-free`: Kilo, LLM7, Routeway, Novita AI, and Ollama Cloud.
+- Added 41 coding-capable API models from those providers, keeping only no-key/free-router models, explicit `:free` zero-price Routeway models, zero-price Novita chat models, and coding-relevant Ollama Cloud models.
+- Added provider metadata, environment-variable support, quota behavior, Settings key-test endpoints, no-key provider handling, and TUI colors for the new providers.
+- Added 💰 billing/credits warnings in TUI Settings, first-run setup, and Web Settings for providers that require paid credits or trial-credit-style accounts despite exposing free/free-tagged models.
+
+### Changed
+- Updated README model/provider counts to 191 models across 20 providers and documented the new providers in the main provider table.
+- Improved first-run setup wording from “Free key” to “Signup/key page” so paid/trial-credit providers are not misleading.
+- Exposed provider billing metadata through the Web Dashboard config API so the desktop/web surface can render the same paid-provider warnings as the TUI.
+
+### Fixed
+- Fixed first-run usability for keyless providers by allowing Kilo and LLM7 to count as usable providers without forcing the API key wizard.
+- Fixed Docker dashboard 404s by serving the Web Router Dashboard aliases (`/api/router/status`, `/api/router/stats`, `/api/router/sets`, `/api/router/tokens`, `/api/router/quick-setup`) and `/api/changelog` directly from the router daemon.
+- Fixed confusing "pending" status for models not in the active router set. Models outside the set now show a distinct "NOT IN SET" label with a dim dot and faded row, instead of the animated yellow "wait" indicator. Applies to Web Dashboard (both Docker daemon and local dev mode).

package/changelog/v0.5.28.md

@@ -0,0 +1,5 @@
+# Changelog v0.5.28 - 2026-06-11
+
+### Fixed
+- Fixed Docker dashboard 404s — the router daemon now serves 10 web dashboard API aliases (`/api/router/status`, `/api/router/stats`, `/api/router/sets`, `/api/router/tokens`, `/api/router/quick-setup`, `/api/router/start`, `/api/router/stop`, `/api/router/probe-mode`, `/api/router/sets/:name/*`, `/api/changelog`) directly, so the React frontend works identically in Docker mode and local dev. Cross-origin guards protect mutating endpoints. (Fixes #116, reported by @stgreenb)
+- Fixed confusing "pending" status for models not in the active router set — models outside the set now show a distinct "NOT IN SET" label with a dim gray dot and faded row opacity, instead of the animated yellow "wait" indicator. The `inRouterSet` field is now exposed by both the Docker daemon and the local web server. (Fixes #117, reported by @stgreenb)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "free-coding-models",
-  "version": "0.5.26",
+  "version": "0.5.28",
   "description": "Find the fastest coding LLM models in seconds — ping free models from multiple providers, pick the best one for OpenCode, Cursor, or any AI coding assistant.",
   "keywords": [
     "nvidia",

package/sources.js

@@ -29,7 +29,7 @@
  *   📖 Secondary: https://swe-rebench.com (independent evals, scores are lower)
  *   📖 Leaderboard tracker: https://www.marc0.dev/en/leaderboard
  *
- *   @exports nvidiaNim, groq, cerebras, sambanova, openrouter, githubModels, mistral, codestral, scaleway, googleai, zai, qwen, cloudflare, ovhcloud, opencodeZen — model arrays per active provider
+ *   @exports nvidiaNim, groq, cerebras, sambanova, openrouter, githubModels, mistral, codestral, scaleway, googleai, zai, qwen, cloudflare, ovhcloud, opencodeZen, kilo, llm7, routeway, novita, ollamaCloud — model arrays per active provider
  *   @exports sources — map of active free/free-limited providers, each with { name, url, models }
 
  *   @exports MODELS — flat array of [modelId, label, tier, sweScore, ctx, providerKey]
@@ -363,6 +363,76 @@ export const opencodeZen = [
   // Added (2026-06-01): minimax-m3-free — MiniMax M3, 1M context, MSA attention (SWE-Bench Pro 59.0%)
 ]
 
+// 📖 Kilo source - https://api.kilo.ai/api/gateway
+// 📖 OpenAI-compatible gateway. `kilo-auto/free` works without a key and routes to Kilo's current free model pool.
+// 📖 Keep only the stable router model here; individual promo `:free` models churn too quickly.
+export const kilo = [
+  ['kilo-auto/free',                         'Kilo Auto Free',      'A+', '-',     '256k'],
+]
+
+// 📖 LLM7 source - https://api.llm7.io/v1
+// 📖 Free unauthenticated tier works with tight shared limits; optional free token at https://token.llm7.io
+// 📖 Pro-tagged models from /v1/models are intentionally excluded.
+export const llm7 = [
+  ['qwen3-235b',                             'Qwen3 235B',          'S+', '70.0%', '240k'],
+  ['mistral-small-3.2',                      'Mistral Small 3.2',   'B+', '34.0%', '128k'],
+  ['codestral-latest',                       'Codestral Latest',    'B+', '34.0%', '32k'],
+  ['devstral-small-2:24b',                   'Devstral Small 2',    'A',  '-',     '128k'],
+]
+
+// 📖 Routeway source - https://api.routeway.ai/v1/models
+// 📖 OpenAI-compatible gateway with explicit zero-price `:free` chat models.
+// 📖 Live catalog checked 2026-06-11; only chat-completions models with free pricing are listed.
+export const routeway = [
+  ['deepseek-v4-flash:free',                 'DeepSeek V4 Flash',   'S+', '72.0%', '1M'],
+  ['step-3.5-flash:free',                    'Step 3.5 Flash',      'S+', '74.4%', '256k'],
+  ['ling-2.6-flash:free',                    'Ling 2.6 Flash',      'S',  '-',     '262k'],
+  ['gpt-oss-120b:free',                      'GPT OSS 120B',        'S',  '60.0%', '131k'],
+  ['minimax-m2:free',                        'MiniMax M2',          'S',  '-',     '197k'],
+  ['laguna-m.1:free',                        'Poolside Laguna M.1', 'S+', '-',     '131k'],
+  ['laguna-xs.2:free',                       'Poolside Laguna XS.2','S+', '-',     '131k'],
+  ['gemma-4-31b-it:free',                    'Gemma 4 31B',         'A',  '45.0%', '262k'],
+  ['nemotron-3-nano-30b-a3b:free',           'Nemotron Nano 30B',   'A',  '43.0%', '256k'],
+  ['llama-3.3-70b-instruct:free',            'Llama 3.3 70B',       'A-', '39.5%', '131k'],
+  ['mistral-nemo-instruct:free',             'Mistral Nemo',        'B+', '30.0%', '16k'],
+  ['nemotron-nano-9b-v2:free',               'Nemotron Nano 9B',    'B+', '18.0%', '128k'],
+  ['llama-3.1-8b-instruct:free',             'Llama 3.1 8B',        'B',  '28.8%', '16k'],
+  ['llama-3.2-3b-instruct:free',             'Llama 3.2 3B',        'B',  '20.0%', '16k'],
+  ['llama-3.2-1b-instruct:free',             'Llama 3.2 1B',        'C',  '-',     '16k'],
+]
+
+// 📖 Novita AI source - https://api.novita.ai/openai/v1/models
+// 📖 Novita is mostly paid/trial-credit, so this catalog only includes live chat models reporting 0 input/output price.
+// 📖 Test/dev/placeholder zero-price IDs were intentionally excluded.
+export const novita = [
+  ['qwen/qwen3.6-plus',                      'Qwen3.6 Plus',        'S+', '72.0%', '1M'],
+  ['qwen/qwen3.5-plus',                      'Qwen3.5 Plus',        'S',  '68.0%', '1M'],
+  ['nex-agi/nex-n2-pro',                     'Nex N2 Pro',          'S',  '-',     '262k'],
+  ['minimax/m2-her',                         'MiniMax M2 HER',      'S',  '-',     '32k'],
+]
+
+// 📖 Ollama Cloud source - https://ollama.com/pricing and https://ollama.com/v1/models
+// 📖 Free plan includes cloud model access with session/weekly limits. This list keeps coding-relevant cloud models only.
+export const ollamaCloud = [
+  ['minimax-m2.7',                           'MiniMax M2.7',        'S+', '80.2%', '200k'],
+  ['glm-5.1',                                'GLM 5.1',             'S+', '77.8%', '203k'],
+  ['kimi-k2.6',                              'Kimi K2.6',           'S+', '76.8%', '131k'],
+  ['deepseek-v4-pro',                        'DeepSeek V4 Pro',     'S+', '73.1%', '1M'],
+  ['deepseek-v4-flash',                      'DeepSeek V4 Flash',   'S+', '72.0%', '1M'],
+  ['devstral-2:123b',                        'Devstral 2 123B',     'S+', '72.2%', '200k'],
+  ['qwen3-coder:480b',                       'Qwen3 Coder 480B',    'S+', '70.6%', '256k'],
+  ['nemotron-3-ultra',                       'Nemotron 3 Ultra',    'S+', '-',     '1M'],
+  ['glm-4.7',                                'GLM 4.7',             'S+', '73.8%', '128k'],
+  ['qwen3.5:397b',                           'Qwen3.5 397B',        'S',  '68.0%', '128k'],
+  ['qwen3-coder-next',                       'Qwen3 Coder Next',    'S',  '65.0%', '256k'],
+  ['gpt-oss:120b',                           'GPT OSS 120B',        'S',  '60.0%', '128k'],
+  ['minimax-m3',                             'MiniMax M3',          'S',  '59.0%', '1M'],
+  ['mistral-large-3:675b',                   'Mistral Large 3',     'A+', '58.0%', '256k'],
+  ['nemotron-3-super',                       'Nemotron 3 Super',    'A+', '56.0%', '128k'],
+  ['gemma4:31b',                             'Gemma 4 31B',         'A',  '45.0%', '256k'],
+  ['gpt-oss:20b',                            'GPT OSS 20B',         'A',  '42.0%', '128k'],
+]
+
 // 📖 All sources combined - used by the main script
 // 📖 Each source has: name (display), url (API endpoint), models (array of model tuples)
 // 📖 Providers ordered by generosity of free tier (most generous first)
@@ -445,6 +515,33 @@ export const sources = {
     models: opencodeZen,
     zenOnly: true,
   },
+  kilo: {
+    name: 'Kilo',
+    url: 'https://api.kilo.ai/api/gateway/chat/completions',
+    models: kilo,
+    noKeyNeeded: true,
+  },
+  llm7: {
+    name: 'LLM7',
+    url: 'https://api.llm7.io/v1/chat/completions',
+    models: llm7,
+    noKeyNeeded: true,
+  },
+  routeway: {
+    name: 'Routeway',
+    url: 'https://api.routeway.ai/v1/chat/completions',
+    models: routeway,
+  },
+  novita: {
+    name: 'Novita AI',
+    url: 'https://api.novita.ai/openai/v1/chat/completions',
+    models: novita,
+  },
+  'ollama-cloud': {
+    name: 'Ollama Cloud',
+    url: 'https://ollama.com/v1/chat/completions',
+    models: ollamaCloud,
+  },
 }
 
 // 📖 Flatten all models from all sources — each entry includes providerKey as 6th element

package/src/core/config.js

@@ -134,6 +134,11 @@ const ENV_VARS = {
   zai:        'ZAI_API_KEY',
   ovhcloud:   'OVH_AI_ENDPOINTS_ACCESS_TOKEN',
   'opencode-zen': 'OPENCODE_ZEN_API_KEY',
+  kilo:       ['KILO_API_KEY', 'KILO_TOKEN'],
+  llm7:       'LLM7_API_KEY',
+  routeway:   'ROUTEWAY_API_KEY',
+  novita:     'NOVITA_API_KEY',
+  'ollama-cloud': 'OLLAMA_API_KEY',
 }
 
 // 📖 Smart Router defaults are intentionally conservative: balanced probing,
@@ -205,6 +210,18 @@ function normalizeFavoriteList(favorites) {
   return normalized
 }
 
+// 📖 normalizeStringSet: Convert an array (from JSON) or existing Set into a Set<string>.
+// 📖 Used for hiddenModels and similar set-shaped config fields.
+function normalizeStringSet(value) {
+  if (value instanceof Set) return value
+  if (!Array.isArray(value)) return new Set()
+  const result = new Set()
+  for (const entry of value) {
+    if (typeof entry === 'string' && entry.trim()) result.add(entry.trim())
+  }
+  return result
+}
+
 function normalizeApiKeyValue(value) {
   if (Array.isArray(value)) {
     const normalized = []
@@ -258,6 +275,7 @@ function normalizeSettingsSection(settings) {
     hideUnconfiguredModels: typeof safeSettings.hideUnconfiguredModels === 'boolean' ? safeSettings.hideUnconfiguredModels : true,
     favoritesPinnedAndSticky: typeof safeSettings.favoritesPinnedAndSticky === 'boolean' ? safeSettings.favoritesPinnedAndSticky : false,
     runAiSpeedTestOnStartup: typeof safeSettings.runAiSpeedTestOnStartup === 'boolean' ? safeSettings.runAiSpeedTestOnStartup : false,
+    autoHideBrokenModels: typeof safeSettings.autoHideBrokenModels === 'boolean' ? safeSettings.autoHideBrokenModels : true,
     theme: ['dark', 'light', 'auto'].includes(safeSettings.theme) ? safeSettings.theme : 'auto',
   }
 }
@@ -466,8 +484,9 @@ function normalizeConfigShape(config) {
     favorites: normalizeFavoriteList(safeConfig.favorites),
     telemetry: normalizeTelemetrySection(safeConfig.telemetry),
     endpointInstalls: normalizeEndpointInstalls(safeConfig.endpointInstalls),
-
-
+    // 📖 hiddenModels: Set of "provider/modelId" keys auto-hidden by the 404 probe (Ctrl+Shift+P).
+    // 📖 Only populated when settings.autoHideBrokenModels is true (default).
+    hiddenModels: normalizeStringSet(safeConfig.hiddenModels),
   }
   const normalizedRouter = normalizeRouterConfig(safeConfig.router)
   if (normalizedRouter) normalized.router = normalizedRouter
@@ -679,8 +698,12 @@ export function saveConfig(config, options = {}) {
 
   try {
     const persistedConfig = buildPersistedConfig(config, readStoredConfigSnapshot(), options)
-    const json = JSON.stringify(persistedConfig, null, 2)
-    writeFileSync(tempPath, json, { mode: 0o600 })
+    // 📖 Serialize Sets to arrays for JSON compatibility (e.g. hiddenModels)
+    const jsonSafe = JSON.stringify(persistedConfig, (key, value) => {
+      if (value instanceof Set) return [...value]
+      return value
+    }, 2)
+    writeFileSync(tempPath, jsonSafe, { mode: 0o600 })
     renameSync(tempPath, CONFIG_PATH)
 
     // 📖 Verify the write succeeded by reading back and validating
@@ -1077,6 +1100,7 @@ export function _emptyProfileSettings() {
     hideUnconfiguredModels: true, // 📖 true = default to providers that are actually configured
     favoritesPinnedAndSticky: false, // 📖 default mode keeps favorites as normal starred rows; press Y to pin+stick them.
     runAiSpeedTestOnStartup: false, // 📖 opt-in: automatically fire the Ctrl+U global AI Speed Test after startup.
+    autoHideBrokenModels: true, // 📖 opt-out: auto-hide models that return 404/410 from probe (Ctrl+Shift+P).
     preferredToolMode: 'opencode', // 📖 remember the last Z-selected launcher across app restarts
     theme: 'auto',        // 📖 'auto' follows the terminal/OS theme, override with 'dark' or 'light' if needed
   }
@@ -1126,5 +1150,6 @@ function _emptyConfig() {
     telemetry: { enabled: null, consentVersion: 0, anonymousId: null },
     endpointInstalls: [],
     settings: _emptyProfileSettings(),
+    hiddenModels: new Set(),
   }
 }

package/src/core/provider-metadata.js

@@ -8,7 +8,8 @@
  *   This module owns three separate concerns that all relate to "knowing about providers":
  *
  *   1. `PROVIDER_METADATA` — human-readable display info (label, colour, signup URL, rate limits)
- *      used in the setup wizard (`promptApiKey`) and the Settings overlay.
+ *      used in the setup wizard (`promptApiKey`) and the Settings overlay. Providers that need
+ *      credits/billing can expose `paidProviderNote`, which Settings renders as a 💰 warning.
  *
  *   2. `ENV_VAR_NAMES` — maps providerKey → the environment variable name that carries the API key.
  *      Used when spawning OpenCode child processes so that keys stored only in
@@ -23,6 +24,7 @@
  *
  * @exports
  *   PROVIDER_METADATA, ENV_VAR_NAMES, OPENCODE_MODEL_MAP,
+ *   getProviderBillingNote, getProviderLabelWithBilling,
  *   isWindows, isMac, isLinux
  *
  * @see bin/free-coding-models.js  — consumes all exports from this module
@@ -55,6 +57,11 @@ export const ENV_VAR_NAMES = {
   ovhcloud:   'OVH_AI_ENDPOINTS_ACCESS_TOKEN',
   qwen:       'DASHSCOPE_API_KEY',
   'opencode-zen': 'OPENCODE_ZEN_API_KEY',
+  kilo:       'KILO_API_KEY',
+  llm7:       'LLM7_API_KEY',
+  routeway:   'ROUTEWAY_API_KEY',
+  novita:     'NOVITA_API_KEY',
+  'ollama-cloud': 'OLLAMA_API_KEY',
 }
 
 // 📖 OPENCODE_MODEL_MAP: sparse table of model IDs that differ between sources.js and OpenCode's
@@ -70,6 +77,7 @@ export const OPENCODE_MODEL_MAP = {
 // 📖 `color` is a chalk function for visual distinction in the TUI.
 // 📖 `signupUrl` / `signupHint` guide users through first-time key generation.
 // 📖 `rateLimits` gives a quick reminder of the free-tier quota without opening a browser.
+// 📖 `paidProviderNote` marks providers that require credits/billing despite exposing free/trial/free-tagged models.
 export const PROVIDER_METADATA = {
   nvidia: {
     label: 'NVIDIA NIM',
@@ -143,6 +151,7 @@ export const PROVIDER_METADATA = {
     signupUrl: 'https://deepinfra.com/login',
     signupHint: 'Login → API keys',
     rateLimits: 'Free tier: 200 concurrent requests (default)',
+    paidProviderNote: 'trial credit provider',
   },
   fireworks: {
     label: 'Fireworks AI',
@@ -157,6 +166,7 @@ export const PROVIDER_METADATA = {
     signupUrl: 'https://console.mistral.ai/api-keys',
     signupHint: 'La Plateforme → API keys (MISTRAL_API_KEY; CODESTRAL_API_KEY also works)',
     rateLimits: 'Codestral free access: 30 req/min, 2000/day',
+    paidProviderNote: 'paid - free Experiment plan',
   },
   hyperbolic: {
     label: 'Hyperbolic',
@@ -192,6 +202,7 @@ export const PROVIDER_METADATA = {
     signupUrl: 'https://api.together.ai/settings/api-keys',
     signupHint: 'Settings → API keys',
     rateLimits: 'Credits/promos vary by account (check console)',
+    paidProviderNote: 'trial credit provider',
   },
   cloudflare: {
     label: 'Cloudflare Workers AI',
@@ -257,4 +268,64 @@ export const PROVIDER_METADATA = {
     signupHint: 'Manager → Public Cloud → AI Endpoints → API keys (optional: sandbox works without key)',
     rateLimits: 'Free sandbox: 2 req/min per IP per model (no key). With API key: 400 RPM',
   },
+  kilo: {
+    label: 'Kilo',
+    color: chalk.rgb(120, 255, 190),
+    signupUrl: 'https://kilo.ai',
+    signupHint: 'No key needed for kilo-auto/free; optional OAuth/API token unlocks more models',
+    rateLimits: 'Free router model works without a key; limits are managed by Kilo',
+    noKeyNeeded: true,
+  },
+  llm7: {
+    label: 'LLM7',
+    color: chalk.rgb(180, 255, 140),
+    signupUrl: 'https://token.llm7.io',
+    signupHint: 'Optional: sign in at token.llm7.io for a free token',
+    rateLimits: 'Free shared tier without key; optional free token improves quota',
+    noKeyNeeded: true,
+  },
+  routeway: {
+    label: 'Routeway',
+    color: chalk.rgb(130, 210, 255),
+    signupUrl: 'https://routeway.ai',
+    signupHint: 'Create account → API key',
+    rateLimits: 'Free :free models with an API key; paid models excluded here',
+    paidProviderNote: 'paid — has :free models',
+  },
+  novita: {
+    label: 'Novita AI',
+    color: chalk.rgb(255, 185, 120),
+    signupUrl: 'https://novita.ai/settings/key-management',
+    signupHint: 'Settings → Key Management → Create API key',
+    rateLimits: 'Only zero-price live chat models are listed; other Novita models are paid/trial-credit',
+    paidProviderNote: 'paid — 3 free models',
+  },
+  'ollama-cloud': {
+    label: 'Ollama Cloud',
+    color: chalk.rgb(230, 230, 230),
+    signupUrl: 'https://ollama.com/settings/keys',
+    signupHint: 'Settings → Keys → Create API key',
+    rateLimits: 'Free plan includes cloud access with session + weekly limits',
+  },
+}
+
+/**
+ * 📖 Return the short paid/billing warning note for a provider, formatted for UI display.
+ * @param {string} providerKey
+ * @returns {string}
+ */
+export function getProviderBillingNote(providerKey) {
+  const note = PROVIDER_METADATA[providerKey]?.paidProviderNote
+  return typeof note === 'string' && note.trim() ? `(${note.trim()})` : ''
+}
+
+/**
+ * 📖 Return a provider label with a small money marker when credits/billing are required.
+ * @param {string} providerKey
+ * @param {string} fallbackLabel
+ * @returns {string}
+ */
+export function getProviderLabelWithBilling(providerKey, fallbackLabel) {
+  const label = PROVIDER_METADATA[providerKey]?.label || fallbackLabel || providerKey
+  return getProviderBillingNote(providerKey) ? `${label} 💰` : label
 }

package/src/core/quota-capabilities.js

@@ -54,6 +54,11 @@ export const PROVIDER_CAPABILITIES = {
   cloudflare: { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'daily' },
   zai: { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'none' },
   'opencode-zen': { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'unknown' },
+  kilo: { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'unknown' },
+  llm7: { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'rolling' },
+  routeway: { telemetryType: 'header', supportsEndpoint: false, usageDisplay: 'percent', resetCadence: 'unknown' },
+  novita: { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'unknown' },
+  'ollama-cloud': { telemetryType: 'unknown', supportsEndpoint: false, usageDisplay: 'ok', resetCadence: 'rolling' },
 }
 
 /** Fallback for unrecognized providers */

package/src/core/router-daemon.js

@@ -49,6 +49,7 @@ import {
 } from './config.js'
 import { buildChatCompletionPingBody, ping, resolveCloudflareUrl, shouldUseDisabledThinkingForProvider } from './ping.js'
 import { benchmarkModel, BENCHMARK_TIMEOUT_MS } from './benchmark.js'
+import { loadChangelog } from './changelog-loader.js'
 import { sendUsageTelemetry } from './telemetry.js'
 
 export const ROUTER_DEFAULT_PORT = 19280
@@ -2397,8 +2398,8 @@ class RouterRuntime {
    * 📖 with only the ones that come back 2xx. Returns the new set + a
    * 📖 sample of probe results so the UI can show "what changed".
    */
-  async handleSyncSetRequest(req, res, requestId) {
-    const url = req.url ? new URL(req.url, 'http://localhost') : null
+  async handleSyncSetRequest(req, res, requestId, routeUrl = null) {
+    const url = routeUrl || (req.url ? new URL(req.url, 'http://localhost') : null)
     const pathname = url ? url.pathname : ''
     const setSyncMatch = pathname.match(/^\/sets\/([^/]+)\/sync$/)
     if (!setSyncMatch) {
@@ -2524,6 +2525,87 @@ class RouterRuntime {
         await this.handleProbeModeRequest(req, res, requestId)
         return
       }
+
+      // 📖 Docker mode serves the built Web Dashboard directly from the daemon
+      // 📖 on :19280. The React app uses the same `/api/router/*` routes as
+      // 📖 local dev (`web/server.js`), so the daemon must expose aliases for
+      // 📖 its canonical `/health`, `/stats`, and `/sets` APIs instead of
+      // 📖 forcing the frontend to special-case Docker.
+      if (req.method === 'GET' && url.pathname === '/api/router/status') {
+        sendJson(res, 200, this.statusPayload(), { 'x-request-id': requestId })
+        return
+      }
+      if (req.method === 'GET' && url.pathname === '/api/router/stats') {
+        sendJson(res, 200, this.statsPayload(), { 'x-request-id': requestId })
+        return
+      }
+      if (req.method === 'GET' && url.pathname === '/api/router/tokens') {
+        sendJson(res, 200, this.tokenTracker.summary(), { 'x-request-id': requestId })
+        return
+      }
+      if (req.method === 'GET' && url.pathname === '/api/router/quick-setup') {
+        const router = this.routerConfig()
+        sendJson(res, 200, {
+          running: true,
+          port: this.port,
+          baseUrl: `http://127.0.0.1:${this.port}/v1`,
+          model: 'fcm',
+          activeSet: router.activeSet || DEFAULT_ROUTER_SETTINGS.activeSet,
+          apiKey: 'not-needed',
+        }, { 'x-request-id': requestId })
+        return
+      }
+      if (url.pathname === '/api/router/start') {
+        if (req.method !== 'POST') {
+          sendError(res, 405, 'Method not allowed', 'invalid_request_error', 'method_not_allowed', requestId, { allowed: ['POST'] })
+          return
+        }
+        if (!isSameOriginOrLocal(req)) {
+          sendError(res, 403, 'Forbidden cross-origin request', 'invalid_request_error', 'forbidden_origin', requestId)
+          return
+        }
+        sendJson(res, 200, { ...this.statusPayload(), alreadyRunning: true }, { 'x-request-id': requestId })
+        return
+      }
+      if (url.pathname === '/api/router/stop') {
+        if (req.method !== 'POST') {
+          sendError(res, 405, 'Method not allowed', 'invalid_request_error', 'method_not_allowed', requestId, { allowed: ['POST'] })
+          return
+        }
+        if (!isSameOriginOrLocal(req)) {
+          sendError(res, 403, 'Forbidden cross-origin request', 'invalid_request_error', 'forbidden_origin', requestId)
+          return
+        }
+        sendJson(res, 200, { ok: true, stopped: true, message: 'Daemon shutting down' }, { 'x-request-id': requestId })
+        setTimeout(() => this.shutdown(0), 50)
+        return
+      }
+      if (url.pathname === '/api/router/probe-mode' && req.method === 'POST') {
+        if (!isSameOriginOrLocal(req)) {
+          sendError(res, 403, 'Forbidden cross-origin request', 'invalid_request_error', 'forbidden_origin', requestId)
+          return
+        }
+        await this.handleProbeModeRequest(req, res, requestId)
+        return
+      }
+      if (req.method === 'GET' && url.pathname === '/api/changelog') {
+        sendJson(res, 200, loadChangelog(), { 'x-request-id': requestId })
+        return
+      }
+      if (url.pathname === '/api/router/sets' || url.pathname.startsWith('/api/router/sets/')) {
+        if (req.method !== 'GET' && !isSameOriginOrLocal(req)) {
+          sendError(res, 403, 'Forbidden cross-origin request', 'invalid_request_error', 'forbidden_origin', requestId)
+          return
+        }
+        const aliasedUrl = new URL(req.url, `http://localhost:${this.port}`)
+        aliasedUrl.pathname = aliasedUrl.pathname.replace(/^\/api\/router/, '')
+        if (/^\/sets\/[^/]+\/sync$/.test(aliasedUrl.pathname) && req.method === 'POST') {
+          await this.handleSyncSetRequest(req, res, requestId, aliasedUrl)
+          return
+        }
+        await this.handleSetsRequest(req, res, aliasedUrl, requestId)
+        return
+      }
       if (url.pathname === '/sets' || url.pathname.startsWith('/sets/')) {
         // 📖 /sets/:name/sync has a different return type (rebuilds the
         // 📖 set from probes) so it gets its own handler.

package/src/core/setup.js

@@ -36,7 +36,7 @@
 import chalk from 'chalk'
 import { createRequire } from 'module'
 import { sources } from '../../sources.js'
-import { PROVIDER_METADATA } from './provider-metadata.js'
+import { PROVIDER_METADATA, getProviderBillingNote, getProviderLabelWithBilling } from './provider-metadata.js'
 import { saveConfig } from './config.js'
 
 const require = createRequire(import.meta.url)
@@ -62,7 +62,8 @@ export async function promptApiKey(config) {
     const meta = PROVIDER_METADATA[key] || {}
     return {
       key,
-      label: meta.label || sources[key]?.name || key,
+      label: getProviderLabelWithBilling(key, sources[key]?.name || key),
+      billingNote: getProviderBillingNote(key),
       color: meta.color || chalk.white,
       url: meta.signupUrl || 'https://example.com',
       hint: meta.signupHint || 'Create API key',
@@ -77,8 +78,8 @@ export async function promptApiKey(config) {
   })
 
   for (const p of providers) {
-    console.log(`  ${p.color('●')} ${chalk.bold(p.label)}`)
-    console.log(chalk.dim(`    Free key at: `) + chalk.cyanBright(p.url))
+    console.log(`  ${p.color('●')} ${chalk.bold(p.label)}${p.billingNote ? ' ' + chalk.yellow(p.billingNote) : ''}`)
+    console.log(chalk.dim(`    Signup/key page: `) + chalk.cyanBright(p.url))
     console.log(chalk.dim(`    ${p.hint}`))
     const answer = await ask(chalk.dim(`  Enter key (or Enter to skip): `))
     console.log()

package/src/tui/app.js

@@ -59,7 +59,7 @@
  *   ⚙️ Configuration:
  *   - API keys stored per-provider in ~/.free-coding-models.json (0600 perms)
  *   - Old ~/.free-coding-models plain-text auto-migrated as nvidia key on first run
- *   - Env vars override config: NVIDIA_API_KEY, GROQ_API_KEY, CEREBRAS_API_KEY, OPENROUTER_API_KEY, GITHUB_TOKEN, MISTRAL_API_KEY, SCALEWAY_API_KEY, GOOGLE_API_KEY, CLOUDFLARE_API_TOKEN, DASHSCOPE_API_KEY, ZAI_API_KEY, etc.
+ *   - Env vars override config: NVIDIA_API_KEY, GROQ_API_KEY, CEREBRAS_API_KEY, OPENROUTER_API_KEY, GITHUB_TOKEN, MISTRAL_API_KEY, SCALEWAY_API_KEY, GOOGLE_API_KEY, CLOUDFLARE_API_TOKEN, DASHSCOPE_API_KEY, ZAI_API_KEY, LLM7_API_KEY, ROUTEWAY_API_KEY, NOVITA_API_KEY, OLLAMA_API_KEY, etc.
  *   - ZAI (z.ai) uses a non-standard base path; cloudflare needs CLOUDFLARE_ACCOUNT_ID in env.
  *   - Cloudflare Workers AI requires both CLOUDFLARE_API_TOKEN (or CLOUDFLARE_API_KEY) and CLOUDFLARE_ACCOUNT_ID
  *   - Models loaded from sources.js — all provider/model definitions are centralized there
@@ -205,10 +205,11 @@ export async function runApp(cliArgs, config, startupOptions = {}) {
 
   // 📖 Profile system removed - API keys now persist permanently across all sessions
 
-  // 📖 Check if any provider has a key — if not, run the first-time setup wizard
-  const hasAnyKey = Object.keys(sources).some(pk => !!getApiKey(config, pk))
+  // 📖 Check if any provider has a key — if not, run the first-time setup wizard.
+  // 📖 Keyless providers (Kilo/LLM7) can still run immediately, so they also count as usable.
+  const hasAnyUsableProvider = Object.keys(sources).some(pk => !!getApiKey(config, pk) || PROVIDER_METADATA[pk]?.noKeyNeeded)
 
-  if (!hasAnyKey) {
+  if (!hasAnyUsableProvider) {
     const result = await promptApiKey(config)
     if (!result) {
       console.log()
@@ -314,6 +315,18 @@ export async function runApp(cliArgs, config, startupOptions = {}) {
       isPinging: false, // 📖 Per-row live flag so Last Ping can keep last value and show a spinner during refresh.
       hidden: false,  // 📖 Simple flag to hide/show models
     }))
+
+  // 📖 Auto-hide models that were previously marked broken by the 404 probe.
+  // 📖 Only applies when autoHideBrokenModels setting is enabled (default).
+  if (config.settings?.autoHideBrokenModels !== false && config.hiddenModels instanceof Set) {
+    for (const r of results) {
+      const modelKey = `${r.providerKey}/${r.modelId}`
+      if (config.hiddenModels.has(modelKey)) {
+        r.hidden = true
+      }
+    }
+  }
+
   syncFavoriteFlags(results, config)
   // 📖 Garbage-collect favorites that reference models no longer in sources.js,
   // 📖 so the router dashboard only shows real, launchable models.
@@ -825,6 +838,10 @@ export async function runApp(cliArgs, config, startupOptions = {}) {
       benchmarkResults: state.benchmarkResults,
       benchmarkRunning: state.benchmarkRunning,
       headerFlashColumn: state.headerFlashColumn,
+      probeRunning: state.probeRunning,
+      probeTotal: state.probeTotal,
+      probeCompleted: state.probeCompleted,
+      probeHiddenCount: state.probeHiddenCount,
     }
     if (state.commandPaletteOpen) {
       if (!state.commandPaletteFrozenTable) {