Add DeepSeek and Qwen Coder provider support#11
Merged
Conversation
Extract the Ollama provider's OpenAI-compatible streaming core into a
reusable OpenAiCompatibleProvider base, then add hosted DeepSeek and Qwen
(DashScope) providers on top of it. Wire them through config (keys, base-URL
overrides, provider inference, defaults), the model catalog (pricing/context/
coding scores for deepseek-v4-{pro,flash} and qwen3-coder-{plus,flash}), and
the cost UI. Update CLI help, .env.example, and README.
https://claude.ai/code/session_01GHiv4kP53a96EWsFEAAqWp
Flip the default priority from performance to balanced so the auto-picked model is cost-aware out of the box (best capability-per-dollar behind a quality floor) rather than most-capable-at-any-price. Update config/catalog tests, CLI help, README, AGENTS.md, and .env.example to match. https://claude.ai/code/session_01GHiv4kP53a96EWsFEAAqWp
Add a /priority slash command to view and switch the cost/performance priority mid-session, re-picking the auto-selected model (unless pinned or governed by local-first routing) via a new AgentLoop.setProvider and a modelPinned config flag. Record on-the-fly provider switching as a follow-up in TODO.md. https://claude.ai/code/session_01GHiv4kP53a96EWsFEAAqWp
- Price the session-end summary from the per-turn accumulated cost (matches /costs) instead of repricing all tokens at the final model's rate, which was wrong after a mid-session /priority model switch. - Validate TINY_CODE_PROVIDER / TINY_CODE_PRIORITY env values; an unrecognized value is now ignored with a warning instead of being cast through and silently mis-picking a model. - Add coverage for the provider-scoped synthetic tool-call id fallback. - Run changeset version: bump to 0.3.0 and generate CHANGELOG. https://claude.ai/code/session_01GHiv4kP53a96EWsFEAAqWp
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Extract the Ollama provider's OpenAI-compatible streaming core into a
reusable OpenAiCompatibleProvider base, then add hosted DeepSeek and Qwen
(DashScope) providers on top of it. Wire them through config (keys, base-URL
overrides, provider inference, defaults), the model catalog (pricing/context/
coding scores for deepseek-v4-{pro,flash} and qwen3-coder-{plus,flash}), and
the cost UI. Update CLI help, .env.example, and README.
https://claude.ai/code/session_01GHiv4kP53a96EWsFEAAqWp