Use Claude Code, Codex, VSCode, Pi, and OpenCode (and 6 other harnesses) for free (1.3B+ free tokens) from your terminal, app, IDE, or phone, and now from the browser with native browser sessions (multi-harness + multi-model) like OpenClaw (voice supported + ToS friendly)
Python
Removed nim setting from .env
Alishahryar145d4dc25d28e
diff --git a/.env.example b/.env.exampleindex fa458d74..07ccf0a3 100644--- a/.env.example+++ b/.env.example@@ -32,26 +32,4 @@ FAST_PREFIX_DETECTION=trueENABLE_NETWORK_PROBE_MOCK=trueENABLE_TITLE_GENERATION_SKIP=trueENABLE_SUGGESTION_MODE_SKIP=true-ENABLE_FILEPATH_EXTRACTION_MOCK=true---# All NVIDIA_NIM_* settings are strictly validated (unknown keys will error).-NVIDIA_NIM_TEMPERATURE=1.0-NVIDIA_NIM_TOP_P=1.0-NVIDIA_NIM_TOP_K=-1-NVIDIA_NIM_MAX_TOKENS=81920-NVIDIA_NIM_PRESENCE_PENALTY=0.0-NVIDIA_NIM_FREQUENCY_PENALTY=0.0-NVIDIA_NIM_MIN_P=0.0-NVIDIA_NIM_REPETITION_PENALTY=1.0-NVIDIA_NIM_SEED=-NVIDIA_NIM_STOP=-NVIDIA_NIM_PARALLEL_TOOL_CALLS=true-NVIDIA_NIM_RETURN_TOKENS_AS_TOKEN_IDS=false-NVIDIA_NIM_INCLUDE_STOP_STR_IN_OUTPUT=false-NVIDIA_NIM_IGNORE_EOS=false-NVIDIA_NIM_MIN_TOKENS=0-NVIDIA_NIM_CHAT_TEMPLATE=""-NVIDIA_NIM_REQUEST_ID=""-NVIDIA_NIM_REASONING_EFFORT=high-NVIDIA_NIM_INCLUDE_REASONING=true\ No newline at end of file+ENABLE_FILEPATH_EXTRACTION_MOCK=true\ No newline at end of filediff --git a/README.md b/README.mdindex 1783077c..6dfc04b1 100644--- a/README.md+++ b/README.md@@ -219,32 +219,6 @@ Browse: [openrouter.ai/models](https://openrouter.ai/models)- **NVIDIA NIM** base URL: `https://integrate.api.nvidia.com/v1`- **OpenRouter** base URL: `https://openrouter.ai/api/v1`-**NIM Settings (prefix `NVIDIA_NIM_`)**--| Variable | Description | Default |-| --------------------------------------- | ----------------------------- | ------- |-| `NVIDIA_NIM_TEMPERATURE` | Sampling temperature | `1.0` |-| `NVIDIA_NIM_TOP_P` | Top-p nucleus sampling | `1.0` |-| `NVIDIA_NIM_TOP_K` | Top-k sampling | `-1` |-| `NVIDIA_NIM_MAX_TOKENS` | Max tokens for generation | `81920` |-| `NVIDIA_NIM_PRESENCE_PENALTY` | Presence penalty | `0.0` |-| `NVIDIA_NIM_FREQUENCY_PENALTY` | Frequency penalty | `0.0` |-| `NVIDIA_NIM_MIN_P` | Min-p sampling | `0.0` |-| `NVIDIA_NIM_REPETITION_PENALTY` | Repetition penalty | `1.0` |-| `NVIDIA_NIM_SEED` | RNG seed (blank = unset) | unset |-| `NVIDIA_NIM_STOP` | Stop string (blank = unset) | unset |-| `NVIDIA_NIM_PARALLEL_TOOL_CALLS` | Parallel tool calls | `true` |-| `NVIDIA_NIM_RETURN_TOKENS_AS_TOKEN_IDS` | Return token ids | `false` |-| `NVIDIA_NIM_INCLUDE_STOP_STR_IN_OUTPUT` | Include stop string in output | `false` |-| `NVIDIA_NIM_IGNORE_EOS` | Ignore EOS token | `false` |-| `NVIDIA_NIM_MIN_TOKENS` | Minimum generated tokens | `0` |-| `NVIDIA_NIM_CHAT_TEMPLATE` | Chat template override | unset |-| `NVIDIA_NIM_REQUEST_ID` | Request id override | unset |-| `NVIDIA_NIM_REASONING_EFFORT` | Reasoning effort | `high` |-| `NVIDIA_NIM_INCLUDE_REASONING` | Include reasoning in response | `true` |--All `NVIDIA_NIM_*` settings are strictly validated; unknown keys with this prefix will cause startup errors.-See [`.env.example`](.env.example) for all supported parameters.## Developmentdiff --git a/config/nim.py b/config/nim.pyindex a95af482..83dd3cd2 100644--- a/config/nim.py+++ b/config/nim.py@@ -1,13 +1,12 @@-"""NVIDIA NIM settings (strict validation)."""+"""NVIDIA NIM settings (fixed values, no env config)."""-from typing import Optional, Literal+from typing import Literal, Optional-from pydantic import Field, field_validator-from pydantic_settings import BaseSettings, SettingsConfigDict+from pydantic import BaseModel, ConfigDict, Field, field_validator-class NimSettings(BaseSettings):- """Strictly validated NVIDIA NIM settings."""+class NimSettings(BaseModel):+ """Fixed NVIDIA NIM settings (not configurable via env)."""temperature: float = Field(1.0, ge=0.0)top_p: float = Field(1.0, ge=0.0, le=1.0)@@ -34,12 +33,7 @@ class NimSettings(BaseSettings):reasoning_effort: Literal["low", "medium", "high"] = "high"include_reasoning: bool = True- model_config = SettingsConfigDict(- env_prefix="NVIDIA_NIM_",- # Rely on global load_dotenv in config.settings to avoid- # reading unrelated .env keys into this settings model.- extra="forbid",- )+ model_config = ConfigDict(extra="forbid")@field_validator("top_k")@classmethod