Condensing Context Ignores Provider API Endpoint Prefix

[strikeoncmputrz]

Plugin Type

VSCode Extension

App Version

v4.140.2

Description

Kilo is working fine in terms of tool calling and completions. My provider base url is set to: https://myhost/v1/

However, when Kilo attempted to condense the context it submits requests to /chat/completions rather than /v1/chat/completions despite my provider base URL explicitly including /v1.

Here are the logged failures:

0.0 s, Process: 57045 cached tokens and 581 new tokens at 145.25 T/s, Generate: 19.73 T/s, Context: 57626 tokens) 
2025-12-24 12:50:41.600 INFO:     192.168.5.2:6534 - "POST /v1/chat/completions HTTP/1.1" 200
2025-12-24 12:50:41.602 INFO:     Received chat completion streaming request a51a7ff9355f447c969b69cae7ce5811
2025-12-24 12:50:48.518 INFO:     Finished chat completion streaming request a51a7ff9355f447c969b69cae7ce5811
2025-12-24 12:50:48.519 INFO:     Metrics (ID: a51a7ff9355f447c969b69cae7ce5811): 62 tokens generated in 6.69 seconds (Queue: 
0.0 s, Process: 57685 cached tokens and 422 new tokens at 119.55 T/s, Generate: 19.63 T/s, Context: 58107 tokens) 
2025-12-24 12:50:49.537 INFO:     192.168.5.2:6534 - "POST /v1/chat/completions HTTP/1.1" 200
2025-12-24 12:50:49.538 INFO:     Received chat completion streaming request c6fcba4bec2f4a488fa6458b75da383d
2025-12-24 12:50:57.386 INFO:     Finished chat completion streaming request c6fcba4bec2f4a488fa6458b75da383d
2025-12-24 12:50:57.387 INFO:     Metrics (ID: c6fcba4bec2f4a488fa6458b75da383d): 98 tokens generated in 7.64 seconds (Queue: 
0.0 s, Process: 58168 cached tokens and 268 new tokens at 99.26 T/s, Generate: 19.85 T/s, Context: 58436 tokens) 
2025-12-24 12:50:58.478 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:50:58.902 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:50:59.369 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:50:59.754 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:00.234 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:00.702 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:01.144 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:01.589 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:02.079 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:02.509 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:02.957 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:03.442 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:03.886 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:04.286 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:04.671 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:05.098 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:05.636 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:06.766 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404
2025-12-24 12:51:07.216 INFO:     192.168.5.2:6534 - "POST /chat/completions HTTP/1.1" 404

Reproduction steps

1. Set BaseURL to openAI compliant API using /v1
2. Wait for Kilo to condense its context

Provider

TabbyAPI

Model

GLM 4.6

System Information

No response