Skip to content

docs(bedrock_embedding): remove unsupported Marengo Embed 3.0 section - #1319

Closed
devin-ai-integration[bot] wants to merge 1 commit into
mainfrom
devin/lit-7267-docs-providersbedrock_embeddingmd-lines-22-33-178-237-506
Closed

devin-ai-integration[bot] wants to merge 1 commit into
mainfrom
devin/lit-7267-docs-providersbedrock_embeddingmd-lines-22-33-178-237-506

Conversation

@devin-ai-integration

Copy link
Copy Markdown
Contributor

Summary

docs/providers/bedrock_embedding.md (added in #1257) documented TwelveLabs Marengo Embed 3.0 (twelvelabs.marengo-embed-3-0-v1:0) as supported, including input_type="text_image" / "multi_input" with media_source / media_sources, inferenceId / bucketOwner passthrough, and 3.0-specific 400 rejection of 2.7-only params.

None of this exists in litellm main:

  • litellm/types/llms/bedrock.py: TWELVELABS_EMBEDDING_INPUT_TYPES = Literal["text", "image", "video", "audio"]
  • litellm/llms/bedrock/embed/twelvelabs_marengo_transformation.py: _transform_request only branches on those four input types; get_supported_openai_params has no media_source / media_sources / inferenceId, so passing them raises UnsupportedParamsError unless drop_params
  • rg 'text_image|multi_input|media_sources|marengo-embed-3' litellm/ has no embedding matches; model_prices_and_context_window.json has no marengo-embed-3 entry

This PR removes the Marengo Embed 3.0 header link, async-invoke table row, input_type parenthetical, the whole ## TwelveLabs Marengo Embed 3.0 section, and the supported-models table row. The 2.7 docs are untouched.

Fixes LIT-7267
https://linear.app/litellm-ai/issue/LIT-7267/docs-providersbedrock-embeddingmd-lines-22-33-178-237-506-marengo

Link to Devin session: https://app.devin.ai/sessions/99f656d9e75848fbaa8e1de063ccec9b
Open in Devin Desktop: https://app.devin.ai/desktop/session/99f656d9e75848fbaa8e1de063ccec9b?variant=devin

…bedding docs

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
@devin-ai-integration

Copy link
Copy Markdown
Contributor Author

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

  • Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
  • Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

  • Disable automatic comment, CI, and merge conflict monitoring

@vercel

vercel Bot commented Sep 8, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated
litellm Ready Ready Preview Sep 8, 2026 8:51pm UTC

Request Review

@devin-ai-integration

Copy link
Copy Markdown
Contributor Author

Live validation for LIT-7267 against litellm main (aa2219d466): this PR's premise no longer holds. Marengo Embed 3.0 landed in BerriAI/litellm on 2026-09-07 (dc09d9e7cf feat(bedrock): add TwelveLabs Marengo Embed 3.0 embeddings, module litellm/llms/bedrock/embed/twelvelabs_marengo_3_transformation.py, model id listed in litellm/constants.py), so the section this PR deletes describes shipped behavior

Proxy from source (uv run litellm --config ..., port 4011) with model: bedrock/us.twelvelabs.marengo-embed-3-0-v1:0, aws_region_name: us-east-1, real AWS credentials, no stubs. Every snippet from the section being removed works:

$ curl localhost:4011/v1/embeddings -d '{"model":"marengo-3","input":"a dog running on the beach","input_type":"text"}'
model=marengo-3 dims=512 first3=[-0.010009765625, -0.08544921875, 0.009033203125]

$ curl ... -d '{"model":"marengo-3","input":"a duck","input_type":"text_image","media_source":"<base64 256x256 png>"}'
model=marengo-3 dims=512 first3=[-0.05810546875, -0.04736328125, 0.018310546875]

$ curl ... -d '{"model":"marengo-3","input":"a photo of <@bird> on water","input_type":"multi_input","media_sources":{"bird":"<base64 png>"}}'
model=marengo-3 dims=512 first3=[-0.061767578125, -0.0771484375, 0.0859375]

The 400 behaviors the section documents also match the code:

$ curl ... -d '{"model":"marengo-3","input":"a duck","input_type":"text_image"}'
HTTP 400  BedrockException - Input type 'text_image' requires the 'media_source' parameter

$ curl ... -d '{"model":"marengo-3","input":"hi","input_type":"text","textTruncate":"end"}'
HTTP 400  BedrockException - Marengo 3.0 does not accept the Marengo 2.7 parameters textTruncate; set drop_params to drop them

For contrast, the same text_image request against bedrock/us.twelvelabs.marengo-embed-2-7-v1:0 is rejected by Bedrock (/inputType: does not have a value in the enumeration ["text", "image"]), which is exactly why the docs split 2.7 and 3.0

The PR was correct on 2026-09-08 when it was opened against the litellm main of that moment, but litellm main has since caught up with the docs. The PR also currently shows a merge conflict with docs main (main moved sk-1234 to placeholders in #1431). Recommendation: close without merging; the current docs/providers/bedrock_embedding.md Marengo 3.0 section is accurate

This branch was successfully deployed

1 active deployment
Preview — 2cbe0a48 Deployed Sep 8, 2026 by vercel[bot]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant