Skip to content
CC Switch WikiUnofficial Community Wiki

There's an issue with the selected model (minimaxai/minimax-m2.5) - CC Switch

Quick fix

For NVIDIA NIM / MiniMax, enable the local proxy and rectifier, or export ANTHROPIC_BASE_URL to the local proxy port.

Symptom

Section titled “Symptom”
Error output
There's an issue with the selected model (minimaxai/minimax-m2.5[1m]). It may not exist or you may not have access to it. Run --model to pick a different model.

Cause

Section titled “Cause”

This error usually occurs when switching through cc-switch to third-party APIs such as NVIDIA NIM or MiniMax. The model mapping may not be loaded correctly into the local proxy, so requests are not forwarded to the correct model endpoint. NVIDIA API also does not support the Anthropic `prompt_cache_key` parameter; if the rectifier is not enabled to filter it, calls can fail. Some users still encounter 502 after manually setting `ANTHROPIC_BASE_URL`, indicating a connectivity or configuration issue between the local proxy and the upstream service.

Fix

Section titled “Fix”

  1. Enable the local proxy and its “intercept Claude” feature so Claude Code requests are forwarded through the cc-switch local proxy.

  2. Enable the rectifier in Advanced Settings to automatically filter the `prompt_cache_key` parameter unsupported by NVIDIA API.

  3. Check the API format setting and use OpenAI chat completions rather than the Anthropic native format.

  4. If the error persists, manually set the environment variable in the terminal before starting Claude Code:

    终端环境变量
    export ANTHROPIC_BASE_URL="http://127.0.0.1:代理端口"

  5. Update cc-switch to the latest version (v3.12.3+ includes multiple compatibility improvements).

Affected Versions

Section titled “Affected Versions”
ToolClaude Code
Version3.11 - 3.12.0
PlatformsWindowsmacOSLinux

Source Issues

Section titled “Source Issues”

This page is distilled from 2 real issues

FAQ

Section titled “FAQ”
Why do I still get the same error after enabling the proxy and interception?
Confirm that the rectifier is enabled and that the model name exists in the current provider's model mapping. NVIDIA's m2.5 model may not support 1m context; try another model.
What should I do if I get 502 after manually exporting ANTHROPIC_BASE_URL?
A 502 error usually means the local proxy cannot connect to the upstream API server. Check whether the proxy port is occupied and whether the NVIDIA/MiniMax API key and base_url configuration are correct.
What base_url should I use for StepFun?
For pay-as-you-go, use `https://api.stepfun.com`; for the Step Plan subscription, use `https://api.stepfun.com/step_plan`. The current cc-switch built-in preset is ambiguous. v3.20.4 separates them in Codex; separation for Claude Code is not yet implemented.
Section titled “Related problems”
Cache hit rate only 40% and extremely slow - Claude CodeDisable the dynamic attribution header in `.claude/settings.json`; if that fails, switch to a model with working caching and monitor the main issue.2 issues mergedProvider 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 mergedModel list is empty; Codex cannot select the DeepSeek modelUpgrade to CC Switch 4.0 and reapply the DeepSeek preset; model list display in the Codex desktop app also depends on its version support.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.