[Bug]: Custom `input_cost_per_token_above_ _tokens` tiers are silently dropped unless N is 128k/200k/272k
![[Bug]: Custom `input_cost_per_token_above_ _tokens` tiers are silently dropped unless N is 128k/200k/272k](https://www.chat-gpts.plus/wp-content/uploads/2026/06/30344-fbe5ac89-768x403.jpg)
用户在使用 litellm.register_model 注册自定义模型时,配置了 input_cost_per_token_above_500k_tokens 之类的分层定价键,但在调用 litellm.cost_calculator.cost_per_token 计算成本时,系统并未使用该分层价


![[Bug]: RPC call to sample_tokens timed out. Qwen3.5-397B-A17B](https://www.chat-gpts.plus/wp-content/uploads/2026/06/35496-ef53c44d-768x403.jpg)


![[Bug]: MCP Gateway returns hardcoded `serverInfo.name = "litellm-mcp-server"` for every `/mcp/{alias}` endpoint](https://www.chat-gpts.plus/wp-content/uploads/2026/06/29800-abdeaae3-768x403.jpg)


![[Question]: Why does reranking need to recalculate the scores for vector search and full-text search?](https://www.chat-gpts.plus/wp-content/uploads/2026/06/7784-f56cea63-768x403.jpg)