HKUDS/DeepTutor · error · LLMAPIError

Cohere API error: {error_text}

Error message

Cohere API error: {error_text}

What it means

_cohere_complete POSTs to the Cohere endpoint and on any non-200 status reads the body and raises LLMAPIError with the server's error message, status code, and provider='cohere'. As with the other backends, the embedded text is the provider's own error body, which usually states the reason (invalid token, model not found, rate limit).

Source

Thrown at deeptutor/services/llm/cloud_provider.py:855

    max_tokens_value = max_tokens if max_tokens is not None else 4096
    temperature_value = temperature if temperature is not None else 0.7
    data: dict[str, object] = {
        "model": model,
        "message": f"{system_prompt}\n\n{prompt}",
        "max_tokens": max_tokens_value,
        "temperature": temperature_value,
    }

    timeout = aiohttp.ClientTimeout(total=120)
    connector = _get_aiohttp_connector()
    async with aiohttp.ClientSession(
        timeout=timeout, connector=connector, trust_env=True
    ) as session:
        async with session.post(url, headers=headers, json=data) as response:
            if response.status != 200:
                error_text = await response.text()
                raise LLMAPIError(
                    f"Cohere API error: {error_text}",
                    status_code=response.status,
                    provider="cohere",
                )

            result = cast(dict[str, object], await response.json())
            text = result.get("text")
            if isinstance(text, str):
                return text
            raise LLMAPIError(
                "Cohere API error: unexpected response payload",
                status_code=response.status,
                provider="cohere",
            )


async def fetch_models(
    base_url: str,

View on GitHub (pinned to 3e82f13042)

Solutions

1. Read e.status_code and the embedded error_text for the server's reason.
2. Fix auth (401), model slug (404), or back off (429).
3. Confirm base_url matches the Cohere API version the code targets.
4. Upgrade model names to current Cohere slugs if deprecated.

Example fix

// before
out = await complete(prompt=p, binding="cohere", model="command", api_key=k)

# after
out = await complete(prompt=p, binding="cohere", model="command-r-plus", api_key=k)

try:
    out = await complete(prompt=p, binding="cohere", model=m, api_key=k)
except LLMAPIError as e:
    if e.status_code == 429:
        await asyncio.sleep(15); retry()
    elif e.status_code == 404:
        update_model_slug()
    else:
        raise

When it happens

Trigger: 401/403 invalid Cohere token; 404 wrong model slug or base_url path; 429 trial-key rate limits; 400 from bad parameters on the generate payload.

Common situations: Trial keys with tight rate limits; deprecated model slugs after Cohere renames (command → command-r); custom gateways exposing Cohere under a different prefix.

Related errors

• Anthropic API error: {error_text}
• Anthropic stream error: {error_text}
• OpenAI API error: {error_text}
• OpenAI stream error: {error_text}
• Cohere API key is missing from the active LLM profile.

AI-assisted analysis of HKUDS/DeepTutor@3e82f13042 (2026-08-27). Data as JSON: /api/errors/db14a4e4587d18d4. Report an issue: GitHub.