docs: add Auxen integration page#2990
Conversation
Adds a Model Provider integration page for Auxen (https://auxen.ai), which serves per-customer dedicated LLM endpoints with an OpenAI- compatible /v1/chat/completions API. The page documents tracing Auxen calls via the existing langfuse.openai drop-in wrapper — no Auxen- specific Langfuse SDK is required. AI agent (Claude) assisted in drafting this content.
|
@Samp2Alex is attempting to deploy a commit to the langfuse Team on Vercel. A member of the Team first needs to authorize it. |
|
|
| --- | ||
| title: Observability for Auxen with Langfuse | ||
| sidebarTitle: Auxen | ||
| logo: /images/integrations/openai_icon.svg |
There was a problem hiding this comment.
Wrong logo — OpenAI icon shown on Auxen page
The logo field points to /images/integrations/openai_icon.svg, so the OpenAI logo will be rendered in the Langfuse docs sidebar and page header for this Auxen integration. Every other provider page uses a provider-specific icon (e.g., deepseek_icon.svg, fireworks_ai_icon.svg). Either add an Auxen-specific SVG under public/images/integrations/ or leave the field empty until one is available.
Prompt To Fix With AI
This is a comment left during a code review.
Path: content/integrations/model-providers/auxen.mdx
Line: 4
Comment:
**Wrong logo — OpenAI icon shown on Auxen page**
The `logo` field points to `/images/integrations/openai_icon.svg`, so the OpenAI logo will be rendered in the Langfuse docs sidebar and page header for this Auxen integration. Every other provider page uses a provider-specific icon (e.g., `deepseek_icon.svg`, `fireworks_ai_icon.svg`). Either add an Auxen-specific SVG under `public/images/integrations/` or leave the field empty until one is available.
How can I resolve this? If you propose a fix, please make it concise.
Adds a Model Provider integration page for Auxen, which serves per-customer dedicated LLM endpoints (Llama 3.1/3.2, Qwen 2.5, Mistral, Gemma 2, Mixtral, Phi-3, Command R) with an OpenAI-compatible `/v1/chat/completions` API.
The page documents tracing Auxen calls via the existing `langfuse.openai` drop-in wrapper — no Auxen-specific Langfuse SDK is required. Each Auxen instance has a per-instance base URL of the form `https://api.auxen.ai/v1/inst_xxx/v1\` and an `auxk_*` bearer token, and the OpenAI client just needs both passed through.
Mirrors the structure of the existing DeepSeek / Together / Fireworks pages and gets added to the Model Providers sidebar in alphabetical position.
Companion PRs (Auxen distribution series)
AI agent (Claude) assisted in drafting this PR.
Greptile Summary
This PR adds a new Model Provider integration page for Auxen and inserts it in the alphabetically correct position in the sidebar
meta.json. The page documents how to trace Auxen's OpenAI-compatible endpoints using Langfuse'slangfuse.openaidrop-in wrapper.auxen.mdxfollows the structure of the DeepSeek / Together AI / Fireworks AI pages and covers setup, chat completion,@observe()grouping, and streaming examples.meta.jsoncorrectly inserts\"auxen\"between\"anthropic\"and\"baseten\".openai_icon.svg) as a placeholder, which will display the wrong branding in the sidebar; a provider-specific icon should be added or the field left blank.Confidence Score: 4/5
Safe to merge after fixing the logo — the OpenAI icon placeholder actively shows wrong branding on the Auxen page.
The docs content and code examples are correct, but the
logofrontmatter points toopenai_icon.svg, so users browsing the Langfuse integrations list will see the OpenAI logo next to the Auxen entry. The missingLearnMorecomponent and unusedobserveimport are minor polish items.content/integrations/model-providers/auxen.mdx — logo field and missing LearnMore footer
Sequence Diagram
sequenceDiagram participant App as User Application participant LF as langfuse.openai (wrapper) participant Auxen as Auxen API participant LFCloud as Langfuse Cloud App->>LF: client.chat.completions.create(model, messages) LF->>Auxen: POST /v1/chat/completions Auxen-->>LF: ChatCompletion response LF-->>App: response (transparent pass-through) LF->>LFCloud: Async trace (prompt, completion, tokens, latency)Prompt To Fix All With AI
Reviews (1): Last reviewed commit: "docs: add Auxen integration page" | Re-trigger Greptile