Add available_providers to model_list.json#407
Merged
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Contributor
Author
This stack of pull requests is managed by Graphite. Learn more about stacking. |
Merged
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 4, 2026
This allows callers to specify which providers a model is available on. If empty, then we fallback to the existing behavior. Related to braintrustdata/braintrust-proxy#407
aswink
force-pushed
the
aswin/03-03-update_model_list.json
branch
from
2314bbf to
81d7d3b
Compare
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
73b5963 to
6f5ca02
Compare
aswink
force-pushed
the
aswin/03-03-update_model_list.json
branch
from
81d7d3b to
8ec80cf
Compare
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
6f5ca02 to
ff13a9a
Compare
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 4, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 4, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
ff13a9a to
d95e060
Compare
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
d95e060 to
8268252
Compare
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
8268252 to
fa0ba59
Compare
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
fa0ba59 to
1633ab8
Compare
aswink
deleted the
aswin/03-03-add_available_providers_to_model_list.json
branch
aswink
restored the
aswin/03-03-add_available_providers_to_model_list.json
branch
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 11, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 11, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 11, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
requested review from
Ankur Goyal (ankrgyl),
Olmo Maldonado (ibolmo) and
Ken Jiang (knjiang)
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1633ab8b65
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "Codex (@codex) review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "Codex (@codex) address that feedback".
Comment on lines
+163
to
+165
| for (const p of ["gemini", "xai", "groq", "together_ai", "cerebras", "mistral", "perplexity", "databricks"]) { | ||
| if (provider === p && modelName.startsWith(p + "/")) { | ||
| return modelName.substring(p.length + 1); |
There was a problem hiding this comment.
Adding "gemini" to the generic prefix-stripping loop makes translateToBraintrust return early for every gemini/* name, so the Gemma-specific remap below never executes. In practice, inputs like provider="gemini" with modelName="gemini/gemma-2-9b-it" now map to gemma-2-9b-it instead of google/gemma-2-9b-it, which is the key format already present in model_list.json; this causes find-missing/update flows to miss existing models and can generate duplicate or stale entries.
Useful? React with 👍 / 👎.
| @@ -9,7 +9,11 @@ | |||
| "displayName": "GPT-5 mini", | |||
| "reasoning": true, | |||
| "max_input_tokens": 272000, | |||
| "max_output_tokens": 128000 | |||
| "max_output_tokens": 128000, | |||
| "available_providers": [ | |||
Collaborator
There was a problem hiding this comment.
I wonder if this would deprecate this:
https://github.com/braintrustdata/braintrust-proxy/blob/main/packages/proxy/schema/index.ts#L223
Contributor
Author
There was a problem hiding this comment.
Yeah, I was thinking that would happen naturally as we migrate from proxy to gateway
Olmo Maldonado (ibolmo)
approved these changes
Mar 11, 2026
Collaborator
Olmo Maldonado (ibolmo)
left a comment
Olmo Maldonado (ibolmo)
left a comment
There was a problem hiding this comment.
i would double check if there's any side effects to adding this field, but this looks reasonable
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 12, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 12, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
1633ab8 to
230f784
Compare
Without this, it's impossible to tell which models are served by which providers, especially programmatically.
aswink
force-pushed
the
aswin/03-03-add_available_providers_to_model_list.json
branch
from
230f784 to
9c4928c
Compare
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 14, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
deleted the
aswin/03-03-add_available_providers_to_model_list.json
branch
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 14, 2026
For now, the router behavior is the same as it was before, but this opens things up for callers to choose providers when multiple are available for a single model. Related to braintrustdata/braintrust-proxy#407
aswink
added a commit
to braintrustdata/lingua
that referenced
this pull request
Mar 14, 2026
Without this, we can only guess which providers can run a model. For example, it's obvious that a vertex format model would run on GCP Vertex, but what about a model like llama3? It could run on any provider. This change adds explicit mappings from model ID -> supported providers. This allows callers to specify which providers a model is available on. If empty, then we fallback to the existing behavior. Also update provider_aliases so it follows the same logic as resolve_providers, particularly that it returns providers in the same order. Depends on braintrustdata/braintrust-proxy#407
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
2 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.
Without this, it's impossible to tell which models are served by which
providers, especially programmatically.