Skip to content
CC Switch WikiUnofficial Community Wiki

X-OpenAI-Internal-Codex-Responses-Lite - Codex error

Quick fix

Comment out the custom provider model_provider setting and return to the built-in openai provider; alternatively, temporarily switch to gpt-5.4.

Symptom

Section titled “Symptom”
Error output
This model is not supported when using X-OpenAI-Internal-Codex-Responses-Lite.

Cause

Section titled “Cause”

The existing issues do not confirm a single root cause. Reports indicate that after enabling unified session history with a custom provider, Codex encounters a model compatibility issue related to X-OpenAI-Internal-Codex-Responses-Lite and the provider identity path. One user temporarily recovered by commenting out the custom model provider setting and returning to the built-in openai provider.

The maintainer stated that 4.0 rewrote how Codex configuration is written. The write engine now maintains model_provider and the provider table used by unified session history. However, without reproduction details for the new version, it is unclear whether 4.0 fixes this error. Another comment suggested temporarily using gpt-5.4.

Fix

Section titled “Fix”

  1. In config.toml, comment out the custom provider model_provider setting so that Codex returns to the built-in openai provider. The reporter of Issue #4873 confirmed this as a temporary fix.

    config.toml
    model_provider = custom
    # model_provider = custom

  2. Restart Codex, confirm that it is no longer running through the custom provider, and retry the original operation.

  3. If the current routing is still required, temporarily switch to gpt-5.4. This suggestion came from the thread and was not confirmed by the original reporter.

  4. Update to CC Switch 4.0 or the latest version and retest. If the error persists, provide the CC Switch and Codex versions, model, the model_provider setting from config.toml with keys removed, and the complete error message.

Affected Versions

Section titled “Affected Versions”
ToolCodex
Version3.16.4（受影响版本）
PlatformsWindowsmacOS

Source Issues

Section titled “Source Issues”

This page is distilled from 2 real issues

FAQ

Section titled “FAQ”
Does this Responses-Lite error still occur after upgrading to CC Switch 4.0?
The existing issues do not confirm this. The maintainer stated that 4.0 rewrote the relevant configuration-writing logic, but still asked users to retest on the latest version and provide complete details.
Why does commenting out model_provider = custom temporarily restore operation?
The assessment in the thread is that this stops use of the custom provider and returns Codex to the built-in openai provider, avoiding a compatibility issue related to Responses-Lite or the provider identity path. The specific root cause remains unconfirmed.
Section titled “Related problems”
Provider API Key Overwritten: After Codex Restart or SwitchUpgrade to cc-switch v4.0.0 or later to prevent provider keys from overwriting each other and restore the official login state.6 issues mergedRequest exceeds the context windowNo verified solution was provided in the original post; this issue was closed as a duplicate related to catalog configuration and automatic compression.1 issues mergedError from provider (Console Go): Upstream request failedCheck whether the model mapping was applied. If the target model is unavailable or the request body contains an unsupported tool_choice or image… (cc-switch)5 issues merged

This is an unofficial community wiki with no affiliation to the cc-switch authors or the project itself. Its content is compiled from the project's public GitHub issues. This site distributes no software.