hsliuping/TradingAgents-CN · error · ValueError

Unsupported LLM provider: {provider}

Error message

Unsupported LLM provider: {provider}

What it means

Raised by create_llm_client when the requested LLM provider string does not match any of the supported providers after lowercasing. The factory only recognizes a fixed set (e.g. 'anthropic', plus other handled branches above line 53); anything else falls through to this ValueError. It exists to fail fast on typos or unsupported backends before client construction.

Source

Thrown at tradingagents/llm_clients/factory.py:53

    provider_lower = normalize_provider_key(provider)
    provider_lower = _PROVIDER_ALIASES.get(provider_lower, provider_lower)

    if provider_lower in _OPENAI_COMPATIBLE:
        from .openai_client import OpenAIClient

        return OpenAIClient(model, base_url, provider=provider_lower, **kwargs)

    if provider_lower == "google":
        from .google_client import GoogleClient

        return GoogleClient(model, base_url, **kwargs)

    if provider_lower == "anthropic":
        from .anthropic_client import AnthropicClient

        return AnthropicClient(model, base_url, **kwargs)

    raise ValueError(f"Unsupported LLM provider: {provider}")

View on GitHub (pinned to 74783e8817)

Solutions

1. Check the factory's if/elif chain above line 53 for the exact supported provider strings and correct the argument to one of them (e.g. 'anthropic').
2. Print/inspect the exact value being passed (including whitespace and case) at the call site or from config before invoking the factory.
3. If you need a new backend, add a matching branch that imports and returns the corresponding client class instead of bypassing the factory.
4. Upgrade the package if the provider was added in a newer release.

Example fix

// before
client = create_llm_client(provider="Anthropic ", model="claude-3-5-sonnet")

# after
client = create_llm_client(provider="anthropic", model="claude-3-5-sonnet")

from tradingagents.llm_clients.factory import create_llm_client
SUPPORTED_PROVIDERS = {"anthropic", "openai", "deepseek", "google", "ollama"}  # mirror factory branches
provider = (provider or "").strip().lower()
if provider not in SUPPORTED_PROVIDERS:
    raise ConfigError(f"bad provider {provider!r}; expected one of {sorted(SUPPORTED_PROVIDERS)}")
client = create_llm_client(provider, model, base_url, **kwargs)

def is_supported_provider(p: str) -> bool:
    """Check against the factory's if/elif branches (see factory.py)."""
    return isinstance(p, str) and p.strip().lower() in {"anthropic", "openai", "deepseek", "google", "ollama"}

try:
    client = create_llm_client(provider, model, base_url, **kwargs)
except ValueError as e:
    if "Unsupported LLM provider" in str(e):
        raise ConfigError(f"Fix provider setting: {provider!r}") from e
    raise

When it happens

Trigger: Calling create_llm_client(model, base_url) or create_llm_client('openai', ...) with a provider string like 'OpenAI ', 'gpt', 'azure-openai', 'googl', or an unimplemented backend — any value whose lowercased form has no matching if-branch in the factory.

Common situations: Typo in the provider name from config/env (LLM_PROVIDER=opanai), passing a model name instead of a provider name, using a provider the installed version doesn't support (azure, bedrock, ollama), or trailing whitespace/case differences if not normalized upstream.

Related errors

• 使用Google需要设置GOOGLE_API_KEY环境变量或在数据库中配置API Key
• 使用SiliconFlow需要设置SILICONFLOW_API_KEY环境变量
• 使用OpenRouter需要设置OPENROUTER_API_KEY或OPENAI_API_KEY环境变量
• 使用AiHubMix需要设置AIHUBMIX_API_KEY环境变量
• 使用Google AI需要在数据库中配置API Key或设置GOOGLE_API_KEY环境变量

AI-assisted analysis of hsliuping/TradingAgents-CN@74783e8817 (2026-08-28). Data as JSON: /api/errors/51f4439aabec1d74. Report an issue: GitHub.