Issue Station
Sign In
Tools hereToolbox

litellm - ✅(Solved) Fix [Bug]: HuggingFace embedding call does not forward extra_headers (e.g. X-HF-Bill-To) [1 pull requests, 1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
BerriAI/litellm#23502•Fetched 2026-04-08 00:44:00
View on GitHub
Comments
0
Participants
1
Timeline
16
Reactions
0
Author
zhongheng-cheng-softbank-co-jpzhongheng-cheng-softbank-co-jp
Participants
zhongheng-cheng-softbank-co-jpzhongheng-cheng-softbank-co-jp
Timeline (top)
referenced ×9cross-referenced ×3labeled ×3closed ×1

Error Message

No error is raised — the headers are silently ignored because the handler defaults

Root Cause

The X-HF-Bill-To header never reaches the HuggingFace API because headers is not passed from main.py:5096 to huggingface_embed.embedding().

Fix Action

Fixed

PR fix notes

PR #23504: fix(huggingface): pass headers to HF embedding call so extra_headers are forwarded

Description (problem / solution / changelog)

Relevant issues

<!-- e.g. "Fixes #000" -->

Fixes #23502

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

  • I have Added testing in the tests/test_litellm/ directory, Adding at least 1 test is a hard requirement - see details
  • My PR passes all unit tests on make test-unit
  • My PR's scope is as isolated as possible, it only solves 1 specific problem
  • I have requested a Greptile review by commenting @greptileai and received a Confidence Score of at least 4/5 before requesting a maintainer review

CI (LiteLLM team)

CI status guideline:

  • 50-55 passing tests: main is stable with minor issues.
  • 45-49 passing tests: acceptable but needs attention
  • <= 40 passing tests: unstable; be careful with your merges and assess the risk.

  • Branch creation CI run
    Link:

  • CI run for the last commit
    Link:

  • Merge / cherry-pick CI run
    Links:

Type

<!-- Select the type of Pull Request --> <!-- Keep only the necessary ones -->

🐛 Bug Fix ✅ Test

Changes

The embedding() function in litellm/main.py accepts a headers parameter containing user-specified extra_headers. All other provider branches (openrouter, vercel_ai_gateway, bedrock, etc.) forward this parameter to their respective handlers, but the huggingface branch does not.

This causes headers like X-HF-Bill-To (used to bill HuggingFace Inference API usage to an organization) to be silently dropped.

The HuggingFace embedding handler (litellm/llms/huggingface/embedding/handler.py) already accepts a headers parameter and merges it with auth headers via validate_environment() — it simply never receives the caller's headers.

Changed files

  • litellm/main.py (modified, +1/-0)
  • tests/test_litellm/llms/huggingface/embedding/test_huggingface_embedding_handler.py (modified, +15/-0)

Code Example

import litellm

# extra_headers should be forwarded to HuggingFace API
response = litellm.embedding(
    model="huggingface/BAAI/bge-small-en-v1.5",
    input=["hello world"],
    extra_headers={"X-HF-Bill-To": "my-org"},  # This header is silently dropped
)

---

No error is raised — the headers are silently ignored because the handler defaults
headers={} and merges with auth headers only.
RAW_BUFFERClick to expand / collapse

Check for existing issues

  • I have searched the existing issues and checked that my issue is not a duplicate.

What happened?

When calling litellm.embedding() with custom_llm_provider="huggingface", the headers parameter from embedding() in litellm/main.py is not passed to huggingface_embed.embedding(). This means any extra_headers set by the user (e.g. X-HF-Bill-To for HuggingFace organization billing) are silently dropped.

Other providers in the same function (e.g. openrouter, vercel_ai_gateway, bedrock) already pass headers correctly. Only the huggingface branch is missing it.

Steps to Reproduce

import litellm

# extra_headers should be forwarded to HuggingFace API
response = litellm.embedding(
    model="huggingface/BAAI/bge-small-en-v1.5",
    input=["hello world"],
    extra_headers={"X-HF-Bill-To": "my-org"},  # This header is silently dropped
)

The X-HF-Bill-To header never reaches the HuggingFace API because headers is not passed from main.py:5096 to huggingface_embed.embedding().

Relevant log output

No error is raised — the headers are silently ignored because the handler defaults
headers={} and merges with auth headers only.

What part of LiteLLM is this about?

SDK (litellm Python package)

What LiteLLM version are you on ?

v1.82.1

Twitter / LinkedIn details

No response

extent analysis

Fix Plan

To fix the issue, we need to pass the headers parameter from litellm.embedding() to huggingface_embed.embedding().

Here are the steps:

  • Modify the litellm.embedding() function in litellm/main.py to pass the headers parameter to huggingface_embed.embedding().
  • Update the huggingface_embed.embedding() function to accept and use the headers parameter.

Code Changes

# In litellm/main.py
def embedding(..., custom_llm_provider="huggingface", ..., headers=None):
    ...
    if custom_llm_provider == "huggingface":
        return huggingface_embed.embedding(..., headers=headers)  # Pass headers here
    ...

# In huggingface_embed.py
def embedding(..., headers=None):
    ...
    # Use the headers parameter when making the API call
    response = requests.post(url, ..., headers=headers)
    ...

Verification

To verify the fix, you can use the following code:

import litellm

response = litellm.embedding(
    model="huggingface/BAAI/bge-small-en-v1.5",
    input=["hello world"],
    extra_headers={"X-HF-Bill-To": "my-org"},
)

# Check if the X-HF-Bill-To header is present in the response
print(response.request.headers)

If the fix is successful, the X-HF-Bill-To header should be present in the response.

Extra Tips

Make sure to test the fix thoroughly to ensure that it works as expected and does not introduce any new issues. Additionally, consider adding logging or monitoring to detect any similar issues in the future.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#ssr#installation#tensor shape#autograd error#training loop#device allocation#model download#tokenizer error

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING