Fosowl/agenticSeek · error · Exception

Together AI response is empty.

Error message

Together AI response is empty.

What it means

together_fn() raises this when the Together client's chat.completions.create returns None, guarding the subsequent response.choices[0].message.content access. Like the Google variant, the raise is re-wrapped by the outer except as 'Together AI API error: Together AI response is empty.'

Source

Thrown at sources/llm_provider.py:321

        except Exception as e:
            raise Exception(f"GOOGLE API error: {str(e)}") from e

    def together_fn(self, history, verbose=False):
        """
        Use together AI for completion
        """
        from together import Together
        client = Together(api_key=self.api_key)
        if self.is_local:
            raise Exception("Together AI is not available for local use. Change config.ini")

        try:
            response = client.chat.completions.create(
                model=self.model,
                messages=history,
            )
            if response is None:
                raise Exception("Together AI response is empty.")
            thought = response.choices[0].message.content
            if verbose:
                print(thought)
            return thought
        except Exception as e:
            raise Exception(f"Together AI API error: {str(e)}") from e

    def deepseek_fn(self, history, verbose=False):
        """
        Use deepseek api to generate text.
        """
        client = OpenAI(api_key=self.api_key, base_url="https://api.deepseek.com")
        if self.is_local:
            raise Exception("Deepseek (API) is not available for local use. Change config.ini")
        try:
            response = client.chat.completions.create(
                model=self.model,
                messages=history,

View on GitHub (pinned to ae57a23577)

Solutions

1. Check network connectivity/proxy settings toward api.together.xyz
2. Verify the model name in config.ini is a valid Together model
3. Upgrade the together SDK and retry
4. Log the raw HTTP response to identify rate limits or service errors

Example fix

// before
response = client.chat.completions.create(model=self.model, messages=history)
thought = response.choices[0].message.content
// after
response = client.chat.completions.create(model=self.model, messages=history)
if not response or not response.choices:
    raise Exception("Together AI response is empty or malformed.")
thought = response.choices[0].message.content or ""

if not cfg.get("together", {}).get("model"):
    raise ValueError("together model missing in config.ini")
if not reachable("https://api.together.xyz"):
    raise ConnectionError("cannot reach Together API")

def is_valid_completion(response) -> bool:
    return (
        response is not None
        and getattr(response, "choices", None)
        and response.choices[0].message is not None
        and response.choices[0].message.content is not None
    )

try:
    thought = provider.together_fn(history)
except Exception as e:
    if "response is empty" in str(e):
        thought = retry_with_backoff(lambda: provider.together_fn(history), retries=3)
    else:
        raise

When it happens

Trigger: client.chat.completions.create(...) against Together's API returns None instead of a completion object — silent SDK/transport failure or an empty response body.

Common situations: Network/proxy interruptions, Together API returning an empty body (rate limiting or service incidents), invalid model name in config.ini, or an outdated together SDK.

Related errors

• Google response is empty.
• OpenAI response is empty.
• Anthropic response is empty.
• Together AI is not available for local use. Change config.in
• Together AI API error: {str(e)}

AI-assisted analysis of Fosowl/agenticSeek@ae57a23577 (2026-08-30). Data as JSON: /api/errors/0da44211313220d1. Report an issue: GitHub.