odysseus-dev/odysseus · error · HTTPException

Cannot reach /v1/models

Error message

Cannot reach /v1/models

What it means

Error "Cannot reach /v1/models" thrown in odysseus-dev/odysseus.

Source

Thrown at routes/session_routes.py:394

            validation_headers = build_headers(effective_api_key, endpoint_base_url or endpoint_url)

        if skip_val:
            # skip_validation = trust the caller and do NOT probe /v1/models.
            # Used for custom endpoints AND for bare placeholder sessions with no
            # model at all (e.g. an email reply draft just needs a session to live
            # in). Probing here was 400-ing those with "Cannot reach /v1/models".
            pass
        elif not model_to_use:
            from src.llm_core import list_model_ids
            ids = list_model_ids(
                endpoint_url,
                timeout=SESSION_MODEL_VALIDATION_TIMEOUT,
                headers=validation_headers,
                owner=user,
                endpoint_id=endpoint_id.strip() if endpoint_id else None,
            )
            if not ids:
                raise HTTPException(400, "Cannot reach /v1/models")
            # Default to the first CHAT model — endpoints often list embedding/
            # tts/whisper models first (e.g. text-embedding-ada-002), which
            # can't hold a conversation.
            _NON_CHAT = ("text-embedding", "embedding", "tts-", "whisper",
                         "text-moderation", "moderation-", "dall-e", "rerank")
            chat_ids = [m for m in ids if not any(p in m.lower() for p in _NON_CHAT)]
            model_to_use = (chat_ids or ids)[0]
        else:
            from src.llm_core import list_model_ids
            import os as _os
            req_base = _os.path.basename(model_to_use.rstrip("/"))
            avail = list_model_ids(
                endpoint_url,
                timeout=SESSION_MODEL_VALIDATION_TIMEOUT,
                headers=validation_headers,
                owner=user,
                endpoint_id=endpoint_id.strip() if endpoint_id else None,
            )

View on GitHub (pinned to f9235ebbf1)

Solutions

1. Verify the endpoint server is running and /v1/models is reachable.
2. Check the base URL and network path, then retry.

When it happens

Trigger: Triggered when the corresponding server-side validation or runtime check at the recorded location rejects the request or operation and returns this error message to the caller.

Common situations: See trigger scenarios.

AI-assisted analysis of odysseus-dev/odysseus@f9235ebbf1 (2026-08-14). Data as JSON: /api/errors/6a85fb94960258bc. Report an issue: GitHub.