perplexity-web-mcp-cli 0.12.2__tar.gz → 0.12.4__tar.gz

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: perplexity-web-mcp-cli
-Version: 0.12.2
+Version: 0.12.4
 Summary: CLI, MCP server, and Anthropic/OpenAI API-compatible interface for Perplexity AI.
 Keywords: perplexity,ai,mcp,anthropic,api,client
 Author: Jacob BD
@@ -62,14 +62,14 @@ Description-Content-Type: text/markdown
 
 MCP server, CLI, and API-compatible interface for Perplexity AI's web interface.
 
-Use your Perplexity Pro/Max subscription to access premium models (GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Super) from the terminal, through MCP tools, or as an API endpoint.
+Use your Perplexity Pro/Max subscription to access premium models (GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Ultra) from the terminal, through MCP tools, or as an API endpoint.
 
 ## Features
 
 - **CLI**: Query Perplexity models directly from the terminal (`pwm ask`, `pwm council`, `pwm research`, `pwm chat`)
 - **MCP Server**: 17 MCP tools for AI agents with citations, rate limit checking, and multi-turn context
 - **API Server**: Drop-in Anthropic Messages API and OpenAI Chat Completions API
-- **6 Models**: GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Super, Sonar 2
+- **6 Models**: GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Ultra, Sonar 2
 - **Thinking Mode**: Extended thinking support for all compatible models
 - **Deep Research**: Full support for Perplexity's Deep Research mode
 - **Multi-Turn Conversations**: State-preserved threaded conversations for both MCP and CLI REPL
@@ -337,7 +337,7 @@ pwm --ai                   # Print comprehensive AI-optimized reference
 | `claude_sonnet` | Anthropic | Toggle | Claude 4.6 Sonnet |
 | `claude_opus` | Anthropic | Toggle | Claude 4.6 Opus (Max tier required) |
 | `gemini_pro` | Google | Always | Gemini 3.1 Pro |
-| `nemotron` | NVIDIA | Always | Nemotron 3 Super 120B |
+| `nemotron` | NVIDIA | Always | Nemotron 3 Ultra 550B |
 
 ### Source Focus
 
@@ -406,7 +406,7 @@ claude mcp add perplexity pwm-mcp
 | `pplx_claude_sonnet` / `pplx_claude_sonnet_think` | Claude 4.6 Sonnet |
 | `pplx_claude_opus` / `pplx_claude_opus_think` | Claude 4.6 Opus (Max tier) |
 | `pplx_gemini_pro_think` | Gemini 3.1 Pro (thinking always on) |
-| `pplx_nemotron_thinking` | Nemotron 3 Super (thinking always on) |
+| `pplx_nemotron_thinking` | Nemotron 3 Ultra (thinking always on) |
 
 **Smart routing (1):**
 
@@ -488,7 +488,7 @@ Our server's `MODEL_MAP` will seamlessly intercept `sonar` (or any other mapped
 | `claude-sonnet-4-6` | Claude 4.6 Sonnet | Toggle |
 | `claude-opus-4-6` | Claude 4.6 Opus | Toggle |
 | `gemini-3.1-pro` | Gemini 3.1 Pro | Always |
-| `nemotron-3-super` / `nemotron` | Nemotron 3 Super | Always |
+| `nemotron-3-ultra` / `nemotron` | Nemotron 3 Ultra | Always |
 
 Legacy aliases (`claude-3-5-sonnet`, `claude-3-opus`) are supported for compatibility.
 

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/README.md

@@ -25,14 +25,14 @@
 
 MCP server, CLI, and API-compatible interface for Perplexity AI's web interface.
 
-Use your Perplexity Pro/Max subscription to access premium models (GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Super) from the terminal, through MCP tools, or as an API endpoint.
+Use your Perplexity Pro/Max subscription to access premium models (GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Ultra) from the terminal, through MCP tools, or as an API endpoint.
 
 ## Features
 
 - **CLI**: Query Perplexity models directly from the terminal (`pwm ask`, `pwm council`, `pwm research`, `pwm chat`)
 - **MCP Server**: 17 MCP tools for AI agents with citations, rate limit checking, and multi-turn context
 - **API Server**: Drop-in Anthropic Messages API and OpenAI Chat Completions API
-- **6 Models**: GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Super, Sonar 2
+- **6 Models**: GPT-5.4, Claude 4.6 Opus, Claude 4.6 Sonnet, Gemini 3.1 Pro, Nemotron 3 Ultra, Sonar 2
 - **Thinking Mode**: Extended thinking support for all compatible models
 - **Deep Research**: Full support for Perplexity's Deep Research mode
 - **Multi-Turn Conversations**: State-preserved threaded conversations for both MCP and CLI REPL
@@ -300,7 +300,7 @@ pwm --ai                   # Print comprehensive AI-optimized reference
 | `claude_sonnet` | Anthropic | Toggle | Claude 4.6 Sonnet |
 | `claude_opus` | Anthropic | Toggle | Claude 4.6 Opus (Max tier required) |
 | `gemini_pro` | Google | Always | Gemini 3.1 Pro |
-| `nemotron` | NVIDIA | Always | Nemotron 3 Super 120B |
+| `nemotron` | NVIDIA | Always | Nemotron 3 Ultra 550B |
 
 ### Source Focus
 
@@ -369,7 +369,7 @@ claude mcp add perplexity pwm-mcp
 | `pplx_claude_sonnet` / `pplx_claude_sonnet_think` | Claude 4.6 Sonnet |
 | `pplx_claude_opus` / `pplx_claude_opus_think` | Claude 4.6 Opus (Max tier) |
 | `pplx_gemini_pro_think` | Gemini 3.1 Pro (thinking always on) |
-| `pplx_nemotron_thinking` | Nemotron 3 Super (thinking always on) |
+| `pplx_nemotron_thinking` | Nemotron 3 Ultra (thinking always on) |
 
 **Smart routing (1):**
 
@@ -451,7 +451,7 @@ Our server's `MODEL_MAP` will seamlessly intercept `sonar` (or any other mapped
 | `claude-sonnet-4-6` | Claude 4.6 Sonnet | Toggle |
 | `claude-opus-4-6` | Claude 4.6 Opus | Toggle |
 | `gemini-3.1-pro` | Gemini 3.1 Pro | Always |
-| `nemotron-3-super` / `nemotron` | Nemotron 3 Super | Always |
+| `nemotron-3-ultra` / `nemotron` | Nemotron 3 Ultra | Always |
 
 Legacy aliases (`claude-3-5-sonnet`, `claude-3-opus`) are supported for compatibility.
 

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "perplexity-web-mcp-cli"
-version = "0.12.2"
+version = "0.12.4"
 description = "CLI, MCP server, and Anthropic/OpenAI API-compatible interface for Perplexity AI."
 authors = [{ name = "Jacob BD" }]
 license = "MIT"

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/api/server.py

@@ -169,12 +169,12 @@ MODEL_MAP: dict[str, tuple[Model, Model | None]] = {
     "gemini-3-pro": (Models.GEMINI_31_PRO_THINKING, Models.GEMINI_31_PRO_THINKING),
     "gemini-pro": (Models.GEMINI_31_PRO_THINKING, Models.GEMINI_31_PRO_THINKING),
     # ==========================================================================
-    # NVIDIA Nemotron 3 Super (via Perplexity)
+    # NVIDIA Nemotron 3 Ultra (via Perplexity)
     # Thinking ALWAYS enabled (no toggle in UI) - reasoning only
     # ==========================================================================
-    "nemotron-3-super": (Models.NEMOTRON_3_SUPER, Models.NEMOTRON_3_SUPER),
-    "nemotron-3": (Models.NEMOTRON_3_SUPER, Models.NEMOTRON_3_SUPER),
-    "nemotron": (Models.NEMOTRON_3_SUPER, Models.NEMOTRON_3_SUPER),
+    "nemotron-3-ultra": (Models.NEMOTRON_3_ULTRA, Models.NEMOTRON_3_ULTRA),
+    "nemotron-3": (Models.NEMOTRON_3_ULTRA, Models.NEMOTRON_3_ULTRA),
+    "nemotron": (Models.NEMOTRON_3_ULTRA, Models.NEMOTRON_3_ULTRA),
     # ==========================================================================
     # Moonshot Kimi Models (via Perplexity)
     # Kimi K2.6 - supports thinking toggle
@@ -198,9 +198,9 @@ AVAILABLE_MODELS = [
     {"id": "gpt-5.5", "description": "GPT-5.5 - OpenAI's latest, Max tier required, thinking toggle available"},
     # Anthropic Claude
     {"id": "claude-sonnet-4-6", "description": "Claude Sonnet 4.6 - Fast, thinking toggle available"},
-    {"id": "claude-opus-4-7", "description": "Claude Opus 4.7 - Advanced reasoning, Max tier required"},
+    {"id": "claude-opus-4-8", "description": "Claude Opus 4.8 - Advanced reasoning, Max tier required"},
     # NVIDIA
-    {"id": "nemotron-3-super", "description": "Nemotron 3 Super - NVIDIA 120B, thinking always on"},
+    {"id": "nemotron-3-ultra", "description": "Nemotron 3 Ultra - NVIDIA 120B, thinking always on"},
     # Moonshot
     {"id": "kimi-k2.6", "description": "Kimi K2.6 - Advanced, thinking toggle available"},
 ]

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/cli/ai_doc.py

@@ -115,9 +115,9 @@ deep_research   pplx_alpha              No         In-depth reports (monthly quo
 gpt54           gpt54                   Yes        OpenAI GPT-5.4 (versatile)
 gpt55           gpt55                   Yes        OpenAI GPT-5.5 (latest, Max tier)
 claude_sonnet   claude46sonnet          Yes        Anthropic Claude 4.6 Sonnet
-claude_opus     claude47opus            Yes        Anthropic Claude 4.7 Opus (Max tier)
+claude_opus     claude48opus            Yes        Anthropic Claude 4.8 Opus (Max tier)
 gemini_pro      gemini31pro_high        Always     Google Gemini 3.1 Pro (thinking only)
-nemotron        nv_nemotron_3_super     Always     NVIDIA Nemotron 3 Super 120B (thinking only)
+nemotron        nv_nemotron_3_ultra     Always     NVIDIA Nemotron 3 Ultra 550B (thinking only)
 kimi_k26        kimi_k26                Yes        Moonshot Kimi K2.6
 
 "Thinking" = extended reasoning mode. Models marked "Always" have thinking
@@ -191,10 +191,10 @@ QUERY TOOLS (each call costs 1 Pro Search query unless noted):
   pplx_gpt55_thinking(query, source_focus="web") GPT-5.5 + thinking — 1 Pro (Max tier)
   pplx_claude_sonnet(query, source_focus="web")   Claude 4.6 Sonnet — 1 Pro
   pplx_claude_sonnet_think(query, source_focus)   Claude 4.6 Sonnet + thinking — 1 Pro
-  pplx_claude_opus(query, source_focus="web")     Claude 4.7 Opus — 1 Pro (Max tier)
-  pplx_claude_opus_think(query, source_focus)     Claude 4.7 Opus + thinking — 1 Pro (Max tier)
+  pplx_claude_opus(query, source_focus="web")     Claude 4.8 Opus — 1 Pro (Max tier)
+  pplx_claude_opus_think(query, source_focus)     Claude 4.8 Opus + thinking — 1 Pro (Max tier)
   pplx_gemini_pro_think(query, source_focus)      Gemini 3.1 Pro (thinking) — 1 Pro
-  pplx_nemotron_thinking(query, source_focus)     Nemotron 3 Super (thinking) — 1 Pro
+  pplx_nemotron_thinking(query, source_focus)     Nemotron 3 Ultra (thinking) — 1 Pro
   pplx_kimi_k26(query, source_focus="web")        Kimi K2.6 — 1 Pro
   pplx_kimi_k26_thinking(query, source_focus)     Kimi K2.6 + thinking — 1 Pro
 

perplexity_web_mcp_cli-0.12.4/src/perplexity_web_mcp/cli/doctor.py

@@ -0,0 +1,208 @@
+"""Diagnostic checks for perplexity-web-mcp installation.
+
+Validates installation, authentication, configuration, rate limits,
+and connectivity -- everything an AI agent or user needs to know
+before using the tool.
+"""
+
+from __future__ import annotations
+
+from importlib import metadata
+import shutil
+
+from perplexity_web_mcp.token_store import TOKEN_FILE, load_token
+
+
+def _check(label: str, ok: bool, detail: str, fix: str = "") -> bool:
+    """Print a check result line and return the status."""
+    icon = "✓" if ok else "✗"
+    line = f"  {icon} {label}: {detail}"
+    if not ok and fix:
+        line += f"  -> {fix}"
+    print(line)
+    return ok
+
+
+def _check_installation() -> bool:
+    print("Installation")
+    all_ok = True
+
+    try:
+        version = metadata.version("perplexity-web-mcp-cli")
+        _check("perplexity-web-mcp-cli", True, version)
+    except metadata.PackageNotFoundError:
+        all_ok = (
+            _check("perplexity-web-mcp-cli", False, "not installed", "pip install perplexity-web-mcp-cli") and all_ok
+        )
+
+    pwm_path = shutil.which("pwm")
+    _check("pwm", pwm_path is not None, pwm_path or "not in PATH", "pip install -e '.[mcp]'")
+
+    pwm_mcp_path = shutil.which("pwm-mcp")
+    all_ok = (
+        _check("pwm-mcp", pwm_mcp_path is not None, pwm_mcp_path or "not in PATH", "pip install -e '.[mcp]'") and all_ok
+    )
+
+    try:
+        import perplexity_web_mcp.api.server  # noqa: F401
+
+        _check("pwm api", True, "module available", "")
+    except ImportError:
+        _check("pwm api", False, "module not found", "pip install -e .")
+
+    return all_ok
+
+
+def _check_authentication(token: str | None, token_exists: bool) -> bool:
+    print("\nAuthentication")
+
+    if not token_exists:
+        return _check("Token", False, "not found", "pwm login")
+
+    token_source = "file" if TOKEN_FILE.exists() else "env"
+    _check("Token", True, f"present ({token_source}, {len(token)} chars)")  # type: ignore[arg-type]
+
+    from perplexity_web_mcp.cli.auth import get_user_info
+
+    user_info = get_user_info(token)
+    if user_info:
+        _check("Account", True, f"{user_info.email}")
+        _check("Subscription", True, user_info.tier_display)
+        return True
+
+    return _check("Account", False, "token invalid or expired", "pwm login")
+
+
+def _check_connectivity(token: str | None, token_exists: bool) -> bool:
+    print("\nConnectivity")
+
+    if not token_exists:
+        _check("Search endpoint", False, "skipped (no token)", "pwm login first")
+        return True
+
+    try:
+        from curl_cffi.requests import Session as CurlSession
+
+        from perplexity_web_mcp.constants import API_BASE_URL, ENDPOINT_SEARCH_INIT, SESSION_COOKIE_NAME
+
+        with CurlSession(impersonate="chrome") as s:
+            s.cookies.set(SESSION_COOKIE_NAME, token)  # type: ignore[arg-type]
+            resp = s.get(f"{API_BASE_URL}{ENDPOINT_SEARCH_INIT}", params={"q": "test"}, timeout=10)
+
+        if resp.status_code == 200:
+            return _check("Search endpoint", True, "reachable")
+        if resp.status_code == 403:
+            return _check(
+                "Search endpoint",
+                False,
+                "403 Forbidden",
+                'Check for VPN/proxy; try: LOG_LEVEL=debug pwm ask "test"',
+            )
+        return _check(
+            "Search endpoint",
+            False,
+            f"unexpected status {resp.status_code}",
+            "check network connection",
+        )
+    except Exception as exc:
+        return _check("Search endpoint", False, f"error: {exc}", "check network connection")
+
+
+def _check_rate_limits(token: str | None, token_exists: bool) -> bool:
+    print("\nRate Limits")
+
+    if not token_exists:
+        _check("Rate limits", False, "skipped (no token)", "pwm login first")
+        return True
+
+    from perplexity_web_mcp.rate_limits import fetch_rate_limits
+
+    limits = fetch_rate_limits(token)
+    if not limits:
+        return _check("Rate limits", False, "could not fetch", "check network connection")
+
+    pro_ok = limits.remaining_pro > 0
+    research_ok = limits.remaining_research > 0
+    all_ok = _check("Pro Search", pro_ok, f"{limits.remaining_pro} remaining", "wait for weekly reset")
+    _check("Deep Research", research_ok, f"{limits.remaining_research} remaining", "wait for monthly reset")
+    _check("Create Files & Apps", True, f"{limits.remaining_labs} remaining")
+    _check("Browser Agent", True, f"{limits.remaining_agentic_research} remaining")
+    return all_ok
+
+
+def _check_mcp_config() -> bool:
+    print("\nMCP Configuration")
+
+    from perplexity_web_mcp.cli.setup import _get_tools, _is_configured_compat
+
+    tools = _get_tools()
+    any_configured = False
+    for tool in tools:
+        configured = _is_configured_compat(tool)
+        if configured:
+            any_configured = True
+        if configured:
+            _check(tool.name, True, "configured")
+        else:
+            _check(tool.name, False, "not configured", f"pwm setup add {tool.name}")
+
+    return any_configured
+
+
+def _check_skills(verbose: bool) -> None:
+    print("\nSkill Installation")
+
+    from perplexity_web_mcp.cli.skill import SKILL_DIR_NAME, _get_current_version, _get_installed_version, _get_targets
+
+    skill_targets = _get_targets()
+    current_ver = _get_current_version()
+    any_skill = False
+    for t in skill_targets:
+        user_ver = _get_installed_version(t.user_dir / SKILL_DIR_NAME)
+        if user_ver:
+            any_skill = True
+            if user_ver == current_ver:
+                _check(t.name, True, f"v{user_ver}")
+            else:
+                _check(t.name, True, f"v{user_ver} (outdated, current: v{current_ver})")
+        elif verbose:
+            _check(t.name, False, "not installed", f"pwm skill install {t.name}")
+
+    if not any_skill:
+        _check("Skills", False, "no skills installed", "pwm skill install claude-code")
+
+
+def _check_security(verbose: bool) -> bool:
+    if not verbose or not TOKEN_FILE.exists():
+        return True
+
+    print("\nSecurity")
+    mode = oct(TOKEN_FILE.stat().st_mode)[-3:]
+    secure = mode == "600"
+    return _check("Token permissions", secure, mode, "chmod 600 ~/.config/perplexity-web-mcp/token")
+
+
+def cmd_doctor(args: list[str]) -> int:
+    """Handle: pwm doctor [--verbose]"""
+    verbose = "-v" in args or "--verbose" in args
+
+    print("\nPerplexity Web MCP Doctor\n")
+
+    token = load_token()
+    token_exists = token is not None and len(token) > 10
+
+    all_ok = _check_installation()
+    all_ok = _check_authentication(token, token_exists) and all_ok
+    all_ok = _check_connectivity(token, token_exists) and all_ok
+    all_ok = _check_rate_limits(token, token_exists) and all_ok
+    all_ok = _check_mcp_config() and all_ok
+    _check_skills(verbose)
+    all_ok = _check_security(verbose) and all_ok
+
+    print()
+    if all_ok:
+        print("✓ All checks passed!")
+    else:
+        print("Some checks failed. See suggestions above (->).")
+
+    return 0 if all_ok else 1

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/cli/main.py

@@ -91,7 +91,7 @@ def cli(ctx):
 # ── Ask ────────────────────────────────────────────────────────────────────
 
 
-@cli.command()
+@cli.command(name="ask")
 @click.argument("query")
 @click.option("-m", "--model", "model_name", default="auto", help=f"Model to use ({', '.join(MODEL_NAMES)}).")
 @click.option("-t", "--thinking", is_flag=True, help="Enable extended thinking mode.")

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/data/SKILL.md

@@ -290,9 +290,9 @@ pwm usage --refresh         # Force-refresh from server
 | `pplx_gpt54` / `_thinking` | 1 Pro | OpenAI GPT-5.4 (versatile) |
 | `pplx_gpt55` / `_thinking` | 1 Pro | OpenAI GPT-5.5 (latest, Max tier) |
 | `pplx_claude_sonnet` / `_think` | 1 Pro | Anthropic Claude 4.6 Sonnet |
-| `pplx_claude_opus` / `_think` | 1 Pro | Anthropic Claude 4.7 Opus |
+| `pplx_claude_opus` / `_think` | 1 Pro | Anthropic Claude 4.8 Opus |
 | `pplx_gemini_pro_think` | 1 Pro | Google Gemini 3.1 Pro (thinking always on) |
-| `pplx_nemotron_thinking` | 1 Pro | NVIDIA Nemotron 3 Super (thinking always on) |
+| `pplx_nemotron_thinking` | 1 Pro | NVIDIA Nemotron 3 Ultra (thinking always on) |
 | `pplx_kimi_k26` / `_thinking` | 1 Pro | Moonshot Kimi K2.6 |
 | `pplx_deep_research` | 1 Research | In-depth reports (**scarce monthly quota**) |
 | `pplx_usage` | FREE | Check remaining quotas |
@@ -317,9 +317,9 @@ For full MCP tool parameters: See [references/mcp-tools.md](references/mcp-tools
 | gpt54 | OpenAI | Toggle | GPT-5.4 (versatile) |
 | gpt55 | OpenAI | Toggle | GPT-5.5 (latest, Max tier) |
 | claude_sonnet | Anthropic | Toggle | Claude 4.6 Sonnet |
-| claude_opus | Anthropic | Toggle | Claude 4.7 Opus (Max tier) |
+| claude_opus | Anthropic | Toggle | Claude 4.8 Opus (Max tier) |
 | gemini_pro | Google | Always | Gemini 3.1 Pro |
-| nemotron | NVIDIA | Always | Nemotron 3 Super 120B |
+| nemotron | NVIDIA | Always | Nemotron 3 Ultra 550B |
 | kimi_k26 | Moonshot | Toggle | Kimi K2.6 |
 
 For full model details: See [references/models.md](references/models.md)

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/data/references/api-endpoints.md

@@ -68,7 +68,7 @@ The API server accepts many model name variants:
 | `gpt-5.4`, `gpt-54`, `gpt54` | GPT-5.4 (versatile) |
 | `gpt-5.5`, `gpt-5-5`, `gpt-55`, `gpt55` | GPT-5.5 (latest, Max tier) |
 | `gemini-3.1-pro`, `gemini-3-pro`, `gemini-pro` | Gemini 3.1 Pro |
-| `nemotron-3-super`, `nemotron` | Nemotron 3 Super |
+| `nemotron-3-ultra`, `nemotron` | Nemotron 3 Ultra |
 
 Unknown model names fall back to `perplexity-auto`.
 

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/data/references/mcp-tools.md

@@ -96,10 +96,10 @@ pplx_<model>(
 | `pplx_gpt55_thinking` | GPT-5.5 (latest, Max tier) | Yes | 1 Pro |
 | `pplx_claude_sonnet` | Claude 4.6 Sonnet | No | 1 Pro |
 | `pplx_claude_sonnet_think` | Claude 4.6 Sonnet | Yes | 1 Pro |
-| `pplx_claude_opus` | Claude 4.7 Opus (Max tier) | No | 1 Pro |
-| `pplx_claude_opus_think` | Claude 4.7 Opus (Max tier) | Yes | 1 Pro |
+| `pplx_claude_opus` | Claude 4.8 Opus (Max tier) | No | 1 Pro |
+| `pplx_claude_opus_think` | Claude 4.8 Opus (Max tier) | Yes | 1 Pro |
 | `pplx_gemini_pro_think` | Gemini 3.1 Pro | Yes (always) | 1 Pro |
-| `pplx_nemotron_thinking` | Nemotron 3 Super | Yes (always) | 1 Pro |
+| `pplx_nemotron_thinking` | Nemotron 3 Ultra | Yes (always) | 1 Pro |
 | `pplx_kimi_k26` | Kimi K2.6 | No | 1 Pro |
 | `pplx_kimi_k26_thinking` | Kimi K2.6 | Yes | 1 Pro |
 

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/data/references/models.md

@@ -45,8 +45,8 @@ Complete list of models available through Perplexity Web MCP.
 - **CLI:** `pwm ask "query" -m claude_sonnet` or `pwm ask "query" -m claude_sonnet -t`
 - **MCP:** `pplx_claude_sonnet(query)` or `pplx_claude_sonnet_think(query)`
 
-### claude_opus (Anthropic Claude 4.7 Opus)
-- **Identifier:** `claude47opus` / `claude47opusthinking`
+### claude_opus (Anthropic Claude 4.8 Opus)
+- **Identifier:** `claude48opus` / `claude48opusthinking`
 - **Thinking:** Toggle
 - **CLI:** `pwm ask "query" -m claude_opus` or `pwm ask "query" -m claude_opus -t`
 - **MCP:** `pplx_claude_opus(query)` or `pplx_claude_opus_think(query)`
@@ -59,12 +59,12 @@ Complete list of models available through Perplexity Web MCP.
 - **MCP:** `pplx_gemini_pro_think(query)` or `pplx_query(query, model="gemini_pro")`
 - **Notes:** Thinking is permanently enabled. The `-t` flag has no effect.
 
-### nemotron (NVIDIA Nemotron 3 Super)
-- **Identifier:** `nv_nemotron_3_super`
+### nemotron (NVIDIA Nemotron 3 Ultra)
+- **Identifier:** `nv_nemotron_3_ultra`
 - **Thinking:** Always on (no non-thinking variant)
 - **CLI:** `pwm ask "query" -m nemotron`
 - **MCP:** `pplx_nemotron_thinking(query)` or `pplx_query(query, model="nemotron")`
-- **Notes:** NVIDIA's Nemotron 3 Super 120B model. Thinking is permanently enabled.
+- **Notes:** NVIDIA's Nemotron 3 Ultra 550B model. Thinking is permanently enabled.
 
 ### kimi_k26 (Moonshot Kimi K2.6)
 - **Identifier:** `kimik26instant` / `kimik26thinking`

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/mcp/server.py

@@ -213,16 +213,16 @@ def pplx_claude_sonnet_think(
 
 @mcp.tool
 def pplx_claude_opus(query: str, source_focus: SourceFocusName = "web", conversation_id: str | None = None) -> str:
-    """Claude Opus 4.7 — Anthropic's most advanced reasoning model. COSTS 1 PRO SEARCH QUERY. Requires Max subscription."""
-    return ask(query, Models.CLAUDE_47_OPUS, source_focus, conversation_id)
+    """Claude Opus 4.8 — Anthropic's most advanced reasoning model. COSTS 1 PRO SEARCH QUERY. Requires Max subscription."""
+    return ask(query, Models.CLAUDE_48_OPUS, source_focus, conversation_id)
 
 
 @mcp.tool
 def pplx_claude_opus_think(
     query: str, source_focus: SourceFocusName = "web", conversation_id: str | None = None
 ) -> str:
-    """Claude Opus 4.7 Thinking — Anthropic's most advanced reasoning model with extended thinking. COSTS 1 PRO SEARCH QUERY. Requires Max subscription."""
-    return ask(query, Models.CLAUDE_47_OPUS_THINKING, source_focus, conversation_id)
+    """Claude Opus 4.8 Thinking — Anthropic's most advanced reasoning model with extended thinking. COSTS 1 PRO SEARCH QUERY. Requires Max subscription."""
+    return ask(query, Models.CLAUDE_48_OPUS_THINKING, source_focus, conversation_id)
 
 
 @mcp.tool
@@ -235,8 +235,8 @@ def pplx_gemini_pro_think(query: str, source_focus: SourceFocusName = "web", con
 def pplx_nemotron_thinking(
     query: str, source_focus: SourceFocusName = "web", conversation_id: str | None = None
 ) -> str:
-    """Nemotron 3 Super — NVIDIA's Nemotron 3 Super 120B model with extended thinking. COSTS 1 PRO SEARCH QUERY."""
-    return ask(query, Models.NEMOTRON_3_SUPER, source_focus, conversation_id)
+    """Nemotron 3 Ultra — NVIDIA's Nemotron 3 Ultra 550B model with extended thinking. COSTS 1 PRO SEARCH QUERY."""
+    return ask(query, Models.NEMOTRON_3_ULTRA, source_focus, conversation_id)
 
 
 @mcp.tool

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/models.py

@@ -55,9 +55,18 @@ class Models:
     CLAUDE_47_OPUS_THINKING = Model(identifier="claude47opusthinking")
     """Claude Opus 4.7 Thinking - Anthropic's most advanced reasoning model (thinking)."""
 
+    CLAUDE_48_OPUS = Model(identifier="claude48opus")
+    """Claude Opus 4.8 - Anthropic's most advanced reasoning model."""
+
+    CLAUDE_48_OPUS_THINKING = Model(identifier="claude48opusthinking")
+    """Claude Opus 4.8 Thinking - Anthropic's most advanced reasoning model (thinking)."""
+
     NEMOTRON_3_SUPER = Model(identifier="nv_nemotron_3_super")
     """Nemotron 3 Super - NVIDIA's Nemotron 3 Super 120B model (thinking)."""
 
+    NEMOTRON_3_ULTRA = Model(identifier="nv_nemotron_3_ultra")
+    """Nemotron 3 Ultra - NVIDIA's Nemotron 3 Ultra 550B model (thinking)."""
+
     KIMI_K2_6 = Model(identifier="kimik26instant")
     """Kimi K2.6 - Moonshot AI's latest model."""
 

{perplexity_web_mcp_cli-0.12.2 → perplexity_web_mcp_cli-0.12.4}/src/perplexity_web_mcp/shared.py

@@ -73,9 +73,9 @@ MODEL_METADATA: dict[str, ModelDefinition] = {
         "Anthropic",
     ),
     "claude_opus": ModelDefinition(
-        Models.CLAUDE_47_OPUS,
-        Models.CLAUDE_47_OPUS_THINKING,
-        "Claude Opus 4.7",
+        Models.CLAUDE_48_OPUS,
+        Models.CLAUDE_48_OPUS_THINKING,
+        "Claude Opus 4.8",
         "Anthropic",
         minimum_tier="max",
     ),
@@ -86,9 +86,9 @@ MODEL_METADATA: dict[str, ModelDefinition] = {
         "Google",
     ),
     "nemotron": ModelDefinition(
-        Models.NEMOTRON_3_SUPER,
-        Models.NEMOTRON_3_SUPER,
-        "Nemotron 3 Super",
+        Models.NEMOTRON_3_ULTRA,
+        Models.NEMOTRON_3_ULTRA,
+        "Nemotron 3 Ultra",
         "NVIDIA",
     ),
     "kimi_k26": ModelDefinition(Models.KIMI_K2_6, Models.KIMI_K2_6_THINKING, "Kimi K2.6", "Moonshot"),
@@ -535,7 +535,7 @@ def council_ask(
     Args:
         query: The question to ask all models.
         models: List of (display_name, Model) tuples. Defaults to
-                GPT-5.4, Claude Opus 4.7, and Gemini 3.1 Pro.
+                GPT-5.4, Claude Opus 4.8, and Gemini 3.1 Pro.
         source_focus: Source focus for all queries.
         synthesize: Whether to run Sonar 2 synthesis (default chairman; still a web query).
         thinking: Use thinking model variants for default council members.

perplexity_web_mcp_cli-0.12.2/src/perplexity_web_mcp/cli/doctor.py

@@ -1,152 +0,0 @@
-"""Diagnostic checks for perplexity-web-mcp installation.
-
-Validates installation, authentication, configuration, rate limits,
-and connectivity -- everything an AI agent or user needs to know
-before using the tool.
-"""
-
-from __future__ import annotations
-
-from importlib import metadata
-import shutil
-
-from perplexity_web_mcp.token_store import TOKEN_FILE, load_token
-
-
-def _check(label: str, ok: bool, detail: str, fix: str = "") -> bool:
-    """Print a check result line and return the status."""
-    icon = "✓" if ok else "✗"
-    line = f"  {icon} {label}: {detail}"
-    if not ok and fix:
-        line += f"  -> {fix}"
-    print(line)
-    return ok
-
-
-def cmd_doctor(args: list[str]) -> int:
-    """Handle: pwm doctor [--verbose]"""
-    verbose = "-v" in args or "--verbose" in args
-    all_ok = True
-
-    print("\nPerplexity Web MCP Doctor\n")
-
-    # --- Installation ---
-    print("Installation")
-    try:
-        version = metadata.version("perplexity-web-mcp-cli")
-        _check("perplexity-web-mcp-cli", True, version)
-    except metadata.PackageNotFoundError:
-        all_ok = (
-            _check("perplexity-web-mcp-cli", False, "not installed", "pip install perplexity-web-mcp-cli") and all_ok
-        )
-
-    pwm_path = shutil.which("pwm")
-    _check("pwm", pwm_path is not None, pwm_path or "not in PATH", "pip install -e '.[mcp]'")
-
-    pwm_mcp_path = shutil.which("pwm-mcp")
-    all_ok = (
-        _check("pwm-mcp", pwm_mcp_path is not None, pwm_mcp_path or "not in PATH", "pip install -e '.[mcp]'") and all_ok
-    )
-
-    try:
-        import perplexity_web_mcp.api.server  # noqa: F401
-
-        _check("pwm api", True, "module available", "")
-    except ImportError:
-        _check("pwm api", False, "module not found", "pip install -e .")
-
-    # --- Authentication ---
-    print("\nAuthentication")
-    token = load_token()
-    token_exists = token is not None and len(token) > 10
-
-    if token_exists:
-        token_source = "file" if TOKEN_FILE.exists() else "env"
-        _check("Token", True, f"present ({token_source}, {len(token)} chars)")
-
-        from perplexity_web_mcp.cli.auth import get_user_info
-
-        user_info = get_user_info(token)
-        if user_info:
-            _check("Account", True, f"{user_info.email}")
-            _check("Subscription", True, user_info.tier_display)
-        else:
-            all_ok = _check("Account", False, "token invalid or expired", "pwm login") and all_ok
-    else:
-        all_ok = _check("Token", False, "not found", "pwm login") and all_ok
-
-    # --- Rate Limits ---
-    print("\nRate Limits")
-    if token_exists:
-        from perplexity_web_mcp.rate_limits import fetch_rate_limits
-
-        limits = fetch_rate_limits(token)
-        if limits:
-            pro_ok = limits.remaining_pro > 0
-            research_ok = limits.remaining_research > 0
-            all_ok = (
-                _check("Pro Search", pro_ok, f"{limits.remaining_pro} remaining", "wait for weekly reset") and all_ok
-            )
-            _check("Deep Research", research_ok, f"{limits.remaining_research} remaining", "wait for monthly reset")
-            _check("Create Files & Apps", True, f"{limits.remaining_labs} remaining")
-            _check("Browser Agent", True, f"{limits.remaining_agentic_research} remaining")
-        else:
-            all_ok = _check("Rate limits", False, "could not fetch", "check network connection") and all_ok
-    else:
-        _check("Rate limits", False, "skipped (no token)", "pwm login first")
-
-    # --- MCP Configuration ---
-    print("\nMCP Configuration")
-    from perplexity_web_mcp.cli.setup import _get_tools, _is_configured_compat
-
-    tools = _get_tools()
-    any_configured = False
-    for tool in tools:
-        configured = _is_configured_compat(tool)
-        if configured:
-            any_configured = True
-        label = tool.name
-        if configured:
-            _check(label, True, "configured")
-        else:
-            _check(label, False, "not configured", f"pwm setup add {tool.name}")
-
-    if not any_configured:
-        all_ok = False
-
-    # --- Skill Installation ---
-    print("\nSkill Installation")
-    from perplexity_web_mcp.cli.skill import SKILL_DIR_NAME, _get_current_version, _get_installed_version, _get_targets
-
-    skill_targets = _get_targets()
-    current_ver = _get_current_version()
-    any_skill = False
-    for t in skill_targets:
-        user_ver = _get_installed_version(t.user_dir / SKILL_DIR_NAME)
-        if user_ver:
-            any_skill = True
-            if user_ver == current_ver:
-                _check(t.name, True, f"v{user_ver}")
-            else:
-                _check(t.name, True, f"v{user_ver} (outdated, current: v{current_ver})")
-        elif verbose:
-            _check(t.name, False, "not installed", f"pwm skill install {t.name}")
-
-    if not any_skill:
-        _check("Skills", False, "no skills installed", "pwm skill install claude-code")
-
-    # --- Token File Permissions ---
-    if verbose and TOKEN_FILE.exists():
-        print("\nSecurity")
-        mode = oct(TOKEN_FILE.stat().st_mode)[-3:]
-        secure = mode == "600"
-        all_ok = _check("Token permissions", secure, mode, "chmod 600 ~/.config/perplexity-web-mcp/token") and all_ok
-
-    # --- Summary ---
-    print()
-    if all_ok:
-        print("✓ All checks passed!")
-    else:
-        print("Some checks failed. See suggestions above (->).")
-
-    return 0 if all_ok else 1