DeepSeek provider — add to settings dropdown #947

New issue

Closed

opened 2026-06-18 10:57:46 +02:00 by sleepy · 0 comments

sleepy commented

2026-06-18 10:57:46 +02:00

Owner

Overview

DeepSeek is already working in our fork, but it requires manual endpoint configuration. Add it as a first-class provider option in the settings dropdown so users can select it without manually configuring base_url.

Upstream References

Upstream has DeepSeek detection in ../odysseus_upstream/src/llm_core.py (host matching api.deepseek.com)
Upstream has dedicated Kimi Code handling: 955455b fix(kimi): resolve Kimi Code API 403 errors and User-Agent restrictions (#3549)
Upstream endpoint resolver handles DeepSeek model listing: ../odysseus_upstream/src/endpoint_resolver.py — api.deepseek.com uses /v1/models endpoint

Implementation

Add DeepSeek as a preset provider in the settings UI dropdown
Auto-fill base_url to https://api.deepseek.com when selected
Auto-fill endpoint paths (chat: /chat/completions, models: /v1/models)
Handle DeepSeek-specific model IDs (DeepSeek-V3, DeepSeek-R1)
Optional: handle Kimi Code User-Agent requirements if using kimi.moonshot.cn

Priority

Low — works already, just UX polish.

## Overview DeepSeek is already working in our fork, but it requires manual endpoint configuration. Add it as a first-class provider option in the settings dropdown so users can select it without manually configuring base_url. ## Upstream References - Upstream has DeepSeek detection in `../odysseus_upstream/src/llm_core.py` (host matching `api.deepseek.com`) - Upstream has dedicated Kimi Code handling: `955455b fix(kimi): resolve Kimi Code API 403 errors and User-Agent restrictions (#3549)` - Upstream endpoint resolver handles DeepSeek model listing: `../odysseus_upstream/src/endpoint_resolver.py` — `api.deepseek.com` uses `/v1/models` endpoint ## Implementation 1. Add DeepSeek as a preset provider in the settings UI dropdown 2. Auto-fill base_url to `https://api.deepseek.com` when selected 3. Auto-fill endpoint paths (chat: `/chat/completions`, models: `/v1/models`) 4. Handle DeepSeek-specific model IDs (DeepSeek-V3, DeepSeek-R1) 5. Optional: handle Kimi Code User-Agent requirements if using kimi.moonshot.cn ## Priority Low — works already, just UX polish.

sleepy referenced this issue from a commit

2026-06-18 18:10:52 +02:00

feat: add DeepSeek provider support to settings dropdown

sleepy referenced this issue from a pull request that will close it,

2026-06-18 18:15:21 +02:00

feat: DeepSeek provider — add to settings dropdown (#947) #958

sleepy closed this issue

2026-06-18 18:15:38 +02:00

sleepy referenced this issue from a commit

2026-06-18 18:32:34 +02:00

Merge PR #958: DeepSeek provider (#947)