utim-cli 2.3.21__tar.gz → 2.3.22__tar.gz

{utim_cli-2.3.21 → utim_cli-2.3.22}/CHANGELOG.md

@@ -1,19 +1,26 @@
 # Changelog
 
+## [2.3.22] - 2026-09-06
+
+### 🛡️ Seamless Rate-Limit Failover to Paid Slugs
+- **Client-Side Auto-Shift**: In `orchestrator_llm_transport.py`, when a free model (e.g. `poolside/laguna-s-2.1:free`) encounters an HTTP 429 or streaming rate-limit event, the client immediately switches to its paid slug (`poolside/laguna-s-2.1`) without terminating the conversation turn or showing an interrupting rate-limit error.
+- **Single Source of Truth**: Centralized `get_paid_slug()` in `utim_cli/constants.py` for uniform mapping across client transport and server completion routing.
+- **Laguna Paid Slug Registration**: Added `poolside/laguna-s-2.1` to the server model registry with complete tool, chat, and reasoning capabilities.
+- **Cost-Protected MiniMax Failover**: Ensured rate-limited MiniMax free models shift to the Laguna S 2.1 paid slug rather than expensive MiniMax paid tiers.
+- **Credit Rate Preservation**: Retained fixed free model pricing ($0.30 input / $0.50 output per 1M tokens) on server fallback turns.
+
 ## [2.3.21] - 2026-09-06
 
 ### 💰 Pricing Modifications
-- **OpenRouter Tier Markup**: Standardized 30% markup across OpenRouter models in the server `model_registry` table.
-- **Free Model Fallback Credit Rate Preservation**: Preserved fixed consumption rates ($0.30 input / $0.50 output per 1M tokens) when transitioning from exhausted APINEX keys to fallback paid slugs, ensuring consistent credit accounting.
+- OpenRouter tier markup standardization and free model fallback credit preservation
 
-### 📦 Models Payload Fixes
-- **Upstream Provider Normalization**: Normalized vendor provider resolution for multi-provider models (MiniMax, Cohere, Dots-Studio, InclusionAI) to route cleanly via OpenRouter upstream.
-- **Image Generation Model Filtering**: Excluded image-generation models from the interactive chat model catalog to avoid completion payload incompatibilities.
-- **APINEX Model Mapping Support**: Added model registry definitions and key rotation support for `qwen/3.8-max` and `free/qwen-3.8-max`.
+### 📦 Models Payload Fixes  
+- Upstream provider normalization and APINEX model mapping support
 
 ### 🛡️ Rate Limit & Vision Tooling Fixes
-- **MiniMax Rate-Limit Fallback to Laguna**: Redirected exhausted MiniMax free models (`minimax/minimax-m2.7:free`, `minimax/minimax-m3:free`) to fall back to `poolside/laguna-s-2.1` paid slug rather than expensive MiniMax paid slugs.
-- **APINEX Transparent Image Analysis**: Updated `read_file` on image files under APINEX models to automatically execute `analyze_image` in the background and return only the text summary, bypassing vision parameter constraints.
+- MiniMax rate-limit fallback to Laguna and APINEX transparent image analysis
+
+*See [docs](docs.md) for detailed APINEX migration information and implementation notes.*
 
 ## [2.3.20] - 2026-09-04
 

{utim_cli-2.3.21/utim_cli.egg-info → utim_cli-2.3.22}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: utim-cli
-Version: 2.3.21
+Version: 2.3.22
 Summary: UTIM – Universal Terminal Intelligence Manager. A powerful agentic AI coding assistant for your terminal.
 License: Emend AI Proprietary EULA
 Project-URL: Homepage, https://utim.dev

{utim_cli-2.3.21 → utim_cli-2.3.22}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "utim-cli"
-version = "2.3.21"
+version = "2.3.22"
 
 description = "UTIM – Universal Terminal Intelligence Manager. A powerful agentic AI coding assistant for your terminal."
 readme = "README.md"

{utim_cli-2.3.21 → utim_cli-2.3.22}/utim/__init__.py

@@ -14,7 +14,7 @@ Quickstart:
 try:
     from utim_cli._version import VERSION as __version__
 except ImportError:
-    __version__ = "2.3.21"
+    __version__ = "2.3.22"
 
 # Developer-friendly core classes (similar to `import genai` / `import openai`)
 from utim_cli.sdk.session import UtimSession as Session

{utim_cli-2.3.21 → utim_cli-2.3.22}/utim_cli/__init__.py

@@ -5,7 +5,7 @@ try:
     from utim_cli._version import VERSION as __version__
 except ImportError:
     # Fallback during first-install / editable-install edge cases
-    __version__ = "2.3.21"
+    __version__ = "2.3.22"
 
 
 import builtins, os, pathlib, urllib.parse, re

{utim_cli-2.3.21 → utim_cli-2.3.22}/utim_cli/_version.py

@@ -11,6 +11,6 @@
 # from here so every surface stays in sync automatically.
 # ─────────────────────────────────────────────────────────────────────────────
 
-VERSION = "2.3.21"
+VERSION = "2.3.22"
 
 

utim_cli-2.3.22/utim_cli/constants.py

@@ -0,0 +1,99 @@
+import os
+
+DEFAULT_MODEL = "poolside/laguna-s-2.1:free"
+SERVER_URL = os.getenv("UTIM_SERVER_URL", "https://api.utim.dev")
+OPENROUTER_BASE_URL = os.getenv("OPENROUTER_BASE_URL", "https://openrouter.ai/api/v1")
+
+SUBAGENT_DEFAULTS = {
+    "web_search": DEFAULT_MODEL,
+    "plan_project": DEFAULT_MODEL,
+    "generate_image": "qwen/qwen-image-3",
+    "analyze_image": "google/gemma-4-31b-it:free",
+}
+
+APINEX_MODEL_IDS = {
+    "claude/opus-5",
+    "claude/sonnet-5",
+    "deepseek/v4-flash-0731",
+    "deepseek/v4-pro-0813",
+    "free/deepseek-v4-flash-0731",
+    "free/deepseek-v4-pro-0813",
+    "free/glm-5.3-flash",
+    "free/gemini-3.1-pro",
+    "free/gemini-3.7-flash",
+    "gemini/3.1-pro",
+    "gemini/3.7-flash",
+    "kimi/k3",
+    "gpt/5.6-luna",
+    "gpt/5.6-sol",
+    "gpt/5.6-terra",
+    "glm/5.3",
+    "glm/5.3-flash",
+    "grok/4.6",
+    "qwen/3.8-max",
+    "free/qwen-3.8-max",
+}
+
+APINEX_TO_OPENROUTER_PAID: dict[str, str] = {
+    "gemini/3.7-flash": "google/gemini-2.5-flash",
+    "free/gemini-3.7-flash": "google/gemini-2.5-flash",
+    "gemini/3.1-pro": "google/gemini-2.5-pro",
+    "free/gemini-3.1-pro": "google/gemini-2.5-pro",
+    "deepseek/v4-flash-0731": "deepseek/deepseek-chat",
+    "free/deepseek-v4-flash-0731": "deepseek/deepseek-chat",
+    "deepseek/v4-pro-0813": "deepseek/deepseek-r1",
+    "free/deepseek-v4-pro-0813": "deepseek/deepseek-r1",
+    "claude/sonnet-5": "anthropic/claude-3.5-sonnet",
+    "claude/opus-5": "anthropic/claude-3-opus",
+    "gpt/5.6-luna": "openai/gpt-4o-mini",
+    "gpt/5.6-sol": "openai/gpt-4o",
+    "gpt/5.6-terra": "openai/gpt-4o",
+    "glm/5.3": "z-ai/glm-4",
+    "glm/5.3-flash": "z-ai/glm-4",
+    "free/glm-5.3-flash": "z-ai/glm-4",
+    "kimi/k3": "moonshotai/kimi-k2",
+    "grok/4.6": "x-ai/grok-2-1212",
+    "qwen/3.8-max": "qwen/qwen-2.5-72b-instruct",
+    "free/qwen-3.8-max": "qwen/qwen-2.5-72b-instruct",
+}
+
+
+def get_paid_slug(model_id: str) -> str:
+    """Return the paid equivalent slug for a free or gateway model.
+    - MiniMax models (e.g. minimax/minimax-m2.7:free, minimax/minimax-m3:free) are pricy,
+      so when hitting rate limits, the server falls back to Laguna S 2.1 paid slug
+      ('poolside/laguna-s-2.1') instead of expensive MiniMax paid slugs.
+    - If model ends with ':free', strip ':free' (provider/model:free -> provider/model).
+    - If APINEX model, map to corresponding OpenRouter paid model.
+    - If model starts with 'free/', check APINEX mapping or strip 'free/'.
+    - If 'openrouter/free', return standard high-capacity default ('google/gemini-2.5-flash').
+    """
+    clean_id = (model_id or "").strip()
+    clean_lower = clean_id.lower()
+
+    # MiniMax free models are pricy on their paid tier; route fallback to Laguna S 2.1 paid slug
+    if "minimax" in clean_lower:
+        return "poolside/laguna-s-2.1"
+
+    if clean_id.endswith(":free"):
+        return clean_id[:-5]
+    if clean_id in APINEX_TO_OPENROUTER_PAID:
+        return APINEX_TO_OPENROUTER_PAID[clean_id]
+    if clean_id.startswith("free/"):
+        unprefixed = clean_id[5:]
+        if unprefixed in APINEX_TO_OPENROUTER_PAID:
+            return APINEX_TO_OPENROUTER_PAID[unprefixed]
+        if unprefixed.endswith(":free"):
+            return unprefixed[:-5]
+        return unprefixed
+    if clean_id == "openrouter/free":
+        return "google/gemini-2.5-flash"
+    return clean_id
+
+
+import sys
+_IS_LEGACY_WIN = (sys.platform == "win32" and "WT_SESSION" not in os.environ)
+PROMPT_SYMBOL = ">" if _IS_LEGACY_WIN else "❯"
+ARROW_SYMBOL = "->" if _IS_LEGACY_WIN else "➔"
+
+

{utim_cli-2.3.21 → utim_cli-2.3.22}/utim_cli/orchestrator_llm_transport.py

@@ -30,7 +30,7 @@ from concurrent.futures import ThreadPoolExecutor, as_completed
 from typing import Any, Dict, List, Optional, Tuple
 
 import requests
-from utim_cli.constants import DEFAULT_MODEL
+from utim_cli.constants import DEFAULT_MODEL, get_paid_slug
 
 # openai SDK removed — we call OpenRouter directly via requests (no Rust/jiter needed)
 from rich.console import Console
@@ -1558,6 +1558,20 @@ class LLMTransportMixin:
                                                     "tool_calls": None,
                                                 }, False
                                             elif _is_rate_limit:
+                                                _paid_slug = get_paid_slug(current_model)
+                                                if _paid_slug != current_model:
+                                                    if not silent:
+                                                        self.console.print(
+                                                            f"\n[bold yellow]⏳  Model '{current_model}' is temporarily rate-limited. Shifting seamlessly to paid slug [bold]{_paid_slug}[/bold]...[/bold yellow]\n"
+                                                        )
+                                                    if _paid_slug not in models_to_try:
+                                                        models_to_try.insert(model_idx + 1, _paid_slug)
+                                                    if not override_model and model_idx == 0:
+                                                        self.model_id = _paid_slug
+                                                        self.model_source = "utim"
+                                                        config.set("main_model_source", "utim")
+                                                    raise RuntimeError(f"__RATE_LIMIT_SHIFT__:{current_model}:{_paid_slug}")
+
                                                 rate_limit_display = (
                                                     f"⏳ **Model Rate Limited**: `{current_model}` is currently experiencing high demand and is temporarily rate-limited by the provider.\n\n"
                                                     f"- **What to do:** Please wait a moment before retrying, or switch to a different model with `/model` or `Ctrl+M`.\n"
@@ -2115,6 +2129,20 @@ class LLMTransportMixin:
                                 "tool_calls": None,
                             }, False
 
+                        _paid_slug = get_paid_slug(current_model)
+                        if _paid_slug != current_model:
+                            if not silent:
+                                self.console.print(
+                                    f"\n[bold yellow]⏳  Model '{current_model}' is temporarily rate-limited. Shifting seamlessly to paid slug [bold]{_paid_slug}[/bold]...[/bold yellow]\n"
+                                )
+                            if _paid_slug not in models_to_try:
+                                models_to_try.insert(model_idx + 1, _paid_slug)
+                            if not override_model and model_idx == 0:
+                                self.model_id = _paid_slug
+                                self.model_source = "utim"
+                                config.set("main_model_source", "utim")
+                            break  # attempt loop → outer loop immediately tries _paid_slug next!
+
                         rate_limit_display = (
                             f"⏳ **Model Rate Limited**: `{current_model}` is currently experiencing high demand and is temporarily rate-limited by the provider.\n\n"
                             f"- **What to do:** Please wait a moment before retrying, or switch to a different model with `/model` or `Ctrl+M`.\n"

{utim_cli-2.3.21 → utim_cli-2.3.22}/utim_cli/server/models.py

@@ -1082,6 +1082,18 @@ MODEL_REGISTRY: dict[MODEL_ID, ModelEntry] = {
         capabilities=["code", "chat", "tool_use"],
         tags=["free"],
     ),
+    "poolside/laguna-s-2.1": ModelEntry(
+        model_id="poolside/laguna-s-2.1",
+        provider="openrouter",
+        cost_input_per_1k=0.000117,
+        cost_output_per_1k=0.000234,
+        cost_input_per_1m=0.117,
+        cost_output_per_1m=0.234,
+        context_window=1_048_576,
+        capabilities=["code", "chat", "tool_use"],
+        tags=["premium", "code"],
+        max_output_tokens=131_072,
+    ),
     "krea/krea-2-medium-turbo": ModelEntry(
         model_id="krea/krea-2-medium-turbo",
         provider="openrouter",

{utim_cli-2.3.21 → utim_cli-2.3.22}/utim_cli/server/routes/completion_routes.py

@@ -556,59 +556,11 @@ def get_openrouter_free_model(model_id: str, db: Optional[Session] = None) -> st
     return free_candidates[0]["model_id"]
 
 
-APINEX_TO_OPENROUTER_PAID: Dict[str, str] = {
-    "gemini/3.7-flash": "google/gemini-2.5-flash",
-    "free/gemini-3.7-flash": "google/gemini-2.5-flash",
-    "gemini/3.1-pro": "google/gemini-2.5-pro",
-    "free/gemini-3.1-pro": "google/gemini-2.5-pro",
-    "deepseek/v4-flash-0731": "deepseek/deepseek-chat",
-    "free/deepseek-v4-flash-0731": "deepseek/deepseek-chat",
-    "deepseek/v4-pro-0813": "deepseek/deepseek-r1",
-    "free/deepseek-v4-pro-0813": "deepseek/deepseek-r1",
-    "claude/sonnet-5": "anthropic/claude-3.5-sonnet",
-    "claude/opus-5": "anthropic/claude-3-opus",
-    "gpt/5.6-luna": "openai/gpt-4o-mini",
-    "gpt/5.6-sol": "openai/gpt-4o",
-    "gpt/5.6-terra": "openai/gpt-4o",
-    "glm/5.3": "z-ai/glm-4",
-    "glm/5.3-flash": "z-ai/glm-4",
-    "free/glm-5.3-flash": "z-ai/glm-4",
-    "kimi/k3": "moonshotai/kimi-k2",
-    "grok/4.6": "x-ai/grok-2-1212",
-}
-
+from utim_cli.constants import (
+    APINEX_TO_OPENROUTER_PAID,
+    get_paid_slug,
+)
 
-def get_paid_slug(model_id: str) -> str:
-    """Return the paid equivalent slug for a free or gateway model.
-    - MiniMax models (e.g. minimax/minimax-m2.7:free, minimax/minimax-m3:free) are pricy,
-      so when hitting rate limits, the server falls back to Laguna S 2.1 paid slug
-      ('poolside/laguna-s-2.1') instead of expensive MiniMax paid slugs.
-    - If model ends with ':free', strip ':free' (provider/model:free -> provider/model).
-    - If APINEX model, map to corresponding OpenRouter paid model.
-    - If model starts with 'free/', check APINEX mapping or strip 'free/'.
-    - If 'openrouter/free', return standard high-capacity default ('google/gemini-2.5-flash').
-    """
-    clean_id = model_id.strip()
-    clean_lower = clean_id.lower()
-
-    # MiniMax free models are pricy on their paid tier; route fallback to Laguna S 2.1 paid slug
-    if "minimax" in clean_lower:
-        return "poolside/laguna-s-2.1"
-
-    if clean_id.endswith(":free"):
-        return clean_id[:-5]
-    if clean_id in APINEX_TO_OPENROUTER_PAID:
-        return APINEX_TO_OPENROUTER_PAID[clean_id]
-    if clean_id.startswith("free/"):
-        unprefixed = clean_id[5:]
-        if unprefixed in APINEX_TO_OPENROUTER_PAID:
-            return APINEX_TO_OPENROUTER_PAID[unprefixed]
-        if unprefixed.endswith(":free"):
-            return unprefixed[:-5]
-        return unprefixed
-    if clean_id == "openrouter/free":
-        return "google/gemini-2.5-flash"
-    return clean_id
 
 
 def is_free_or_gateway_model(model_id: str) -> bool:

{utim_cli-2.3.21 → utim_cli-2.3.22/utim_cli.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: utim-cli
-Version: 2.3.21
+Version: 2.3.22
 Summary: UTIM – Universal Terminal Intelligence Manager. A powerful agentic AI coding assistant for your terminal.
 License: Emend AI Proprietary EULA
 Project-URL: Homepage, https://utim.dev

utim_cli-2.3.21/utim_cli/constants.py

@@ -1,43 +0,0 @@
-import os
-
-DEFAULT_MODEL = "poolside/laguna-s-2.1:free"
-SERVER_URL = os.getenv("UTIM_SERVER_URL", "https://api.utim.dev")
-OPENROUTER_BASE_URL = os.getenv("OPENROUTER_BASE_URL", "https://openrouter.ai/api/v1")
-
-SUBAGENT_DEFAULTS = {
-    "web_search": DEFAULT_MODEL,
-    "plan_project": DEFAULT_MODEL,
-    "generate_image": "qwen/qwen-image-3",
-    "analyze_image": "google/gemma-4-31b-it:free",
-}
-
-APINEX_MODEL_IDS = {
-    "claude/opus-5",
-    "claude/sonnet-5",
-    "deepseek/v4-flash-0731",
-    "deepseek/v4-pro-0813",
-    "free/deepseek-v4-flash-0731",
-    "free/deepseek-v4-pro-0813",
-    "free/glm-5.3-flash",
-    "free/gemini-3.1-pro",
-    "free/gemini-3.7-flash",
-    "gemini/3.1-pro",
-    "gemini/3.7-flash",
-    "kimi/k3",
-    "gpt/5.6-luna",
-    "gpt/5.6-sol",
-    "gpt/5.6-terra",
-    "glm/5.3",
-    "glm/5.3-flash",
-    "grok/4.6",
-    "qwen/3.8-max",
-    "free/qwen-3.8-max",
-}
-
-
-import sys
-_IS_LEGACY_WIN = (sys.platform == "win32" and "WT_SESSION" not in os.environ)
-PROMPT_SYMBOL = ">" if _IS_LEGACY_WIN else "❯"
-ARROW_SYMBOL = "->" if _IS_LEGACY_WIN else "➔"
-
-