xtekky/gpt4free · error · RuntimeError

Failed to chat

Error message

Failed to chat: {response.status} {error_text}

What it means

Raised by DeepAI's chat generator when the POST to 'https://api.deepai.org/hacking_is_a_serious_crime' returns a non-2xx status. This is the main completion request of the provider; the message includes the HTTP status code and response body text, which typically indicates rate limiting or rejection of the generated api key.

Solutions

1. Check the status code in the message: 429 means back off and retry later; 401/403 means the generated api key/headers were rejected.
2. Reduce request frequency or route through a different proxy/IP.
3. Update g4f — DeepAI's key generation recipe (generate_api_key) changes often and is fixed upstream.
4. As a fallback, switch to another image-capable provider in g4f while DeepAI is blocked.

Example fix

# before
for i in range(50):
    out = await DeepAI.create_async_generator(model, messages)

# after
for i in range(50):
    try:
        out = await DeepAI.create_async_generator(model, messages)
    except RuntimeError as e:
        if ' 429 ' in str(e):
            await asyncio.sleep(30)
            continue
        raise

try:
    async for chunk in DeepAI.create_async_generator(model, messages):
        ...
except RuntimeError as e:
    msg = str(e)
    if ' 429 ' in msg:
        await asyncio.sleep(30)  # backoff, then retry
    else:
        raise  # 401/403: generated api key rejected -> update g4f or switch provider

When it happens

Trigger: Calling DeepAI chat (text or after attachment upload) when the endpoint responds 4xx/5xx — commonly 429 rate limit per api key/IP, or 401/403 when the locally generated api_key header is deemed invalid.

Common situations: Bursty traffic from one IP generating many random api keys; DeepAI tightening validation of the hacked api key scheme (provider breaks frequently); proxy or impersonation issues causing WAF blocks; outdated g4f with a stale key recipe.

Related errors

• Cannot create GLM chat session
• Failed to decode JSON from PhindAi response
• Failed to upload file
• TOKEN_ERROR
• Cannot validate GLM account

AI-assisted analysis of xtekky/gpt4free@973504e177 (2026-08-14). Data as JSON: /api/errors/241ef17d7891ed53. Report an issue: GitHub.

Appendix: source

Thrown at g4f/Provider/DeepAI.py:203

                "enabled_tools": json.dumps(
                    ["image_generator", "image_editor"], separators=(",", ":")
                ),
            }

            if attachment_uuids:
                data_dict["attachment_uuids"] = json.dumps(attachment_uuids)

            data = FormData(data_dict)

            async with session.post(
                "https://api.deepai.org/hacking_is_a_serious_crime",
                headers=headers,
                data=data,
                proxy=proxy,
            ) as response:
                if not response.ok:
                    error_text = await response.text()
                    raise RuntimeError(
                        f"Failed to chat: {response.status} {error_text}"
                    )

                buffer = ""
                async for chunk in response.content.iter_any():
                    if chunk:
                        chunk_text = chunk.decode(errors="ignore")
                        if "\x1c" in chunk_text or "\x1c" in buffer:
                            buffer += chunk_text
                        else:
                            yield chunk_text

                if not buffer:
                    return

                parts = buffer.split("\x1c")
                if parts[0].strip():
                    yield parts[0]

View on GitHub (pinned to 973504e177)