update check_ollama_model_exists using ollama.list api.

[GdoongMathew]

Summary

refactor check_ollama_model_exists using ollama provided list api.

[gemini-code-assist]

Code Review

This pull request refactors the check_ollama_model_exists function in api/ollama_patch.py to use the official ollama Python client instead of making raw HTTP requests with requests. The review feedback highlights that the new exact-match logic breaks the original tag-agnostic model matching behavior, and that the exception handling catches ConnectionError instead of the httpx.RequestError and ollama.ResponseError exceptions raised by the underlying client. Suggestions have been provided to restore the tag-agnostic matching and to import httpx to properly handle connection and API errors.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

The new matching logic performs an exact match (model_name == model.model), which breaks the original tag-agnostic matching behavior. For example, if a user checks for llama3 but the installed model is llama3:latest, the check will incorrectly return False.\n\nTo preserve the original behavior, we should compare the base names (by splitting on :).

Suggested change

	is_available = any(model_name == model.model for model in ret.models)
	if is_available:
	logger.info("Ollama model '%s' is available", model_name)
	else:
	logger.warning(f"Could not check Ollama models, status code: {response.status_code}")
	return False
	except requests.exceptions.RequestException as e:
	logger.warning(
	"Ollama model '%s' is not available. Available models: %s. ",
	model_name,
	str([model.model for model in ret.models]))
	model_base = model_name.split(':')[0]
	available_models = [model.model for model in ret.models]
	is_available = any(model_base == m.split(':')[0] for m in available_models)
	if is_available:
	logger.info("Ollama model '%s' is available", model_name)
	else:
	logger.warning(
	"Ollama model '%s' is not available. Available models: %s.",
	model_name,
	str(available_models))

[gemini-code-assist]

The ollama library uses httpx under the hood, which raises httpx.RequestError (not the built-in ConnectionError) for connection issues and timeouts. Additionally, API errors raise ollama.ResponseError. Catching ConnectionError will fail to intercept these exceptions, causing them to fall through to the generic except Exception block.\n\nWe should catch httpx.RequestError and ollama.ResponseError instead.

Suggested change

	except ConnectionError as e:
	logger.warning(f"Could not connect to Ollama to check models: {e}")
	return False
	except (httpx.RequestError, ollama.ResponseError) as e:
	logger.warning(f"Could not connect to Ollama to check models: {e}")
	return False

[GdoongMathew]

Since the timeout argument is set, the actual exception should be httpx.ConnectTimeout instead.

fixed in da71585

[gemini-code-assist]

Import httpx alongside ollama to properly catch connection and request errors raised by the underlying HTTP client.

    import ollama
    import httpx

[GdoongMathew]

added in 592523e