feat(together-ai/google/gemma-4-12B-it): add new models [bot]

[models-bot]

Auto-generated by model-addition-agent for together-ai/google/gemma-4-12B-it.

---

Note

Low Risk
Metadata-only addition of a new model YAML; zero pricing could misreport cost until updated but does not change runtime code paths.

Overview
Adds a new Together AI provider entry for google/gemma-4-12B-it, registering it as a serverless chat model with a 262144 token context window.

Token costs are set to 0.0 with an inline note that the model is not yet listed in Together’s official serverless/pricing docs, so billing metadata may need a follow-up once pricing is published. The file is slimmer than sibling Gemma 4 YAMLs (no status, sources, modalities, or capability flags).

^{Reviewed by Cursor Bugbot for commit 0cad9e9. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Missing multimodal capabilities

Medium Severity

The new config omits modalities, thinking, and features that sibling Gemma 4 models declare. Official docs describe Gemma 4 12B as multimodal (text, image, audio, and video) with thinking support, so registry consumers will under-report its real capabilities.

 

^{Reviewed by Cursor Bugbot for commit 139fb46. Configure here.}

[github-actions]

/test-models

[harshiv-26]

Gateway test results

• Total: 2
• Passed: 0
• Failed: 2
• Validation failed: 0
• Errored: 0
• Skipped: 0
• Success rate: 0.0%

Provider	Model	Scenarios
together-ai	google/gemma-4-12B-it	failure: params, params:stream

Failures (2)

together-ai/google/gemma-4-12B-it — params (failure)

Error

Traceback (most recent call last):
  File "/tmp/tmp4mb1fvxz/snippet.py", line 5, in <module>
    response = client.chat.completions.create(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/_utils/_utils.py", line 286, in wrapper
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/resources/chat/completions/completions.py", line 1147, in create
    return self._post(
           ^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 1259, in post
    return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 1047, in request
    raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'status': 'failure', 'message': 'together-ai error: Unable to access non-serverless model google/gemma-4-12B-it. Please visit https://api.together.ai/models/google/gemma-4-12B-it to create and start a new dedicated endpoint for the model.', 'error': {'message': 'together-ai error: Unable to access non-serverless model google/gemma-4-12B-it. Please visit https://api.together.ai/models/google/gemma-4-12B-it to create and start a new dedicated endpoint for the model.', 'type': 'APIError', 'code': '400'}, 'error_origin_level': 'api_error', 'provider': 'together-ai'}

Code snippet

from openai import OpenAI

client = OpenAI(api_key="***", base_url="https://internal.devtest.truefoundry.tech/api/llm")

response = client.chat.completions.create(
    model="test-v2-together-ai/google-gemma-4-12B-it",
    messages=[
        {"role": "user", "content": "What is the capital of France?"},
    ],
    stream=False,
)

print(response.choices[0].message.content)

together-ai/google/gemma-4-12B-it — params:stream (failure)

Error

Traceback (most recent call last):
  File "/tmp/tmpnqaqz_gv/snippet.py", line 5, in <module>
    response = client.chat.completions.create(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/_utils/_utils.py", line 286, in wrapper
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/resources/chat/completions/completions.py", line 1147, in create
    return self._post(
           ^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 1259, in post
    return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 1047, in request
    raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'status': 'failure', 'message': 'together-ai error: Unable to access non-serverless model google/gemma-4-12B-it. Please visit https://api.together.ai/models/google/gemma-4-12B-it to create and start a new dedicated endpoint for the model.', 'error': {'message': 'together-ai error: Unable to access non-serverless model google/gemma-4-12B-it. Please visit https://api.together.ai/models/google/gemma-4-12B-it to create and start a new dedicated endpoint for the model.', 'type': 'APIError', 'code': '400'}, 'error_origin_level': 'api_error', 'provider': 'together-ai'}

Code snippet

from openai import OpenAI

client = OpenAI(api_key="***", base_url="https://internal.devtest.truefoundry.tech/api/llm")

response = client.chat.completions.create(
    model="test-v2-together-ai/google-gemma-4-12B-it",
    messages=[
        {"role": "user", "content": "What is the capital of France?"},
    ],
    stream=True,
)

for chunk in response:
    if chunk.choices and len(chunk.choices) > 0:
        delta = chunk.choices[0].delta
        if delta.content is not None:
            print(delta.content, end="", flush=True)

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

There are 2 total unresolved issues (including 1 from previous review).

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 0cad9e9. Configure here.}

[cursor]

Nonexistent serverless model entry

High Severity

The file notes that google/gemma-4-12B-it is absent from Together AI docs, yet sets provisioning: serverless. Together AI’s catalog currently lists Gemma 4 31B, not 12B, so this registry entry advertises a serverless model that API callers cannot reach.

 

^{Reviewed by Cursor Bugbot for commit 0cad9e9. Configure here.}