[GH-ISSUE #198] Codex not able to see the model list #70

New issue

Open

opened 2026-10-26 09:44:58 +00:00 by bsakaguchi · 2 comments

bsakaguchi commented

2026-10-26 09:44:58 +00:00

Owner

Originally created by @chingjustwe on GitHub (Jun 8, 2026).
Original GitHub issue: https://github.com/deepseek-ai/awesome-deepseek-agent/issues/198

Note: already tried the FAQ listed on Codex cannot see the model

Originally created by @chingjustwe on GitHub (Jun 8, 2026). Original GitHub issue: https://github.com/deepseek-ai/awesome-deepseek-agent/issues/198 Note: already tried the FAQ listed on [Codex cannot see the model](https://github.com/deepseek-ai/awesome-deepseek-agent/blob/main/docs/codex.md#troubleshooting) <img width="1642" height="958" alt="Image" src="https://github.com/user-attachments/assets/122f9449-26f6-472c-872f-58812d149d6f" />

bsakaguchi commented

2026-10-26 09:44:59 +00:00

Author

Owner

@chingjustwe commented on GitHub (Jun 8, 2026):

Raise PR-199 to fix

@chingjustwe commented on GitHub (Jun 8, 2026): Raise [PR-199](https://github.com/deepseek-ai/awesome-deepseek-agent/pull/199) to fix

bsakaguchi commented

2026-10-26 09:45:00 +00:00

Author

Owner

@tianhao8687 commented on GitHub (Aug 17, 2026):

The current official integration avoids remote model discovery by giving Codex a local model catalog. The most reliable fix now is to follow the current Codex integration guide and regenerate the configuration.

For a manual install, verify that these fields all point to the same catalog and provider:

model = "deepseek-v4-flash"
model_provider = "deepseek"
model_catalog_json = "~/.codex/models.json"

[model_providers.deepseek]
base_url = "https://api.deepseek.com/"
wire_api = "responses"

Also check the following:

~/.codex/models.json really exists under the same home directory used by the running Codex client.
Codex is at least 0.144.0, which is the minimum client version declared by the current catalog.
Fully quit and restart Codex/ChatGPT after changing the files.
On Windows, the desktop model picker may show Custom rather than the DeepSeek display name; the startup banner/model field is the better verification.

If it still fails, the useful diagnostics are codex --version, the resolved catalog path, and the startup banner. Do not post the API key.

@tianhao8687 commented on GitHub (Aug 17, 2026): The current official integration avoids remote model discovery by giving Codex a local model catalog. The most reliable fix now is to follow the current [Codex integration guide](https://api-docs.deepseek.com/quick_start/agent_integrations/codex) and regenerate the configuration. For a manual install, verify that these fields all point to the same catalog and provider: ```toml model = "deepseek-v4-flash" model_provider = "deepseek" model_catalog_json = "~/.codex/models.json" [model_providers.deepseek] base_url = "https://api.deepseek.com/" wire_api = "responses" ``` Also check the following: 1. `~/.codex/models.json` really exists under the same home directory used by the running Codex client. 2. Codex is at least `0.144.0`, which is the minimum client version declared by the current catalog. 3. Fully quit and restart Codex/ChatGPT after changing the files. 4. On Windows, the desktop model picker may show **Custom** rather than the DeepSeek display name; the startup banner/model field is the better verification. If it still fails, the useful diagnostics are `codex --version`, the resolved catalog path, and the startup banner. Do not post the API key.

No labels

pull-request

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

deepseek-ai/awesome-deepseek-agent#70

No description provided.

Rows
Columns