Skip to content

fix: OpenAI-compatible provider calls wrong endpoint (/responses), breaking self-hosted chat backends #63

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
kartikmanimuthu merged 2 commits into from
Oct 19, 2026
Merged

fix: OpenAI-compatible provider calls wrong endpoint (/responses), breaking self-hosted chat backends #63

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5a279cb
fix: openai-compatible provider calls /responses, breaking self-hoste…
fabicarvano Oct 17, 2026
c466248
fix: update openai-compatible mock to match the .chat() routing fix
fabicarvano Oct 17, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion apps/workers/Dockerfile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,7 +51,7 @@ COPY libs/ai/ libs/ai/
RUN cd libs/ai && bunx tsc --project tsconfig.lib.json && \
mkdir -p /app/node_modules/@chatbot/ai && \
cp -rL /app/dist/libs/ai/libs/ai/src /app/node_modules/@chatbot/ai/src && \
printf '{"name":"@chatbot/ai","version":"0.0.1","type":"module","main":"src/index.js","types":"src/index.d.ts"}\n' > /app/node_modules/@chatbot/ai/package.json
printf '{"name":"@chatbot/ai","version":"0.0.1","type":"module","main":"src/index.js","types":"src/index.d.ts","exports":{".":"./src/index.js","./*":"./src/*.js"}}\n' > /app/node_modules/@chatbot/ai/package.json

COPY libs/knowledge-base/ libs/knowledge-base/
RUN cd libs/knowledge-base && bunx tsc --project tsconfig.lib.json && \
Expand Down
6 changes: 6 additions & 0 deletions libs/ai/src/providers/openai-compatible.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,12 @@ import { describe, it, expect, vi, beforeEach } from 'vitest';
const { mockCreateOpenAI, mockStreamText } = vi.hoisted(() => ({
mockCreateOpenAI: vi.fn(() => {
const provider = (model: string) => ({ modelId: model, provider: 'openai-compatible' });
// streamChat calls `this.client.chat(model)`, not the bare callable — the
// bare call routes to OpenAIResponsesModelId/`/v1/responses`, which most
// OpenAI-compatible gateways (Ollama, vLLM, LiteLLM, ...) never implement.
// `.chat()` routes to OpenAIChatModelId/`/v1/chat/completions`, which they
// all support. See openai-compatible.ts's streamChat for the full story.
provider.chat = (model: string) => ({ modelId: model, provider: 'openai-compatible' });
provider.textEmbeddingModel = (model: string) => ({ modelId: model, type: 'embedding' });
return provider;
}),
Expand Down
2 changes: 1 addition & 1 deletion libs/ai/src/providers/openai-compatible.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,7 +41,7 @@ export class OpenAICompatibleProvider implements LLMProvider {
onFinish,
} = options;
return streamText({
model: this.client(model ?? this.chatModel),
model: this.client.chat(model ?? this.chatModel),
messages,
system,
temperature,
Expand Down