feat(together-ai): update model YAMLs [bot]

[models-bot]

Auto-generated by poc-agent for provider together-ai.

---

Note

Low Risk
Metadata-only catalog changes with no application code; the main user impact is routing away from a newly retired model and updated cost/modality metadata.

Overview
Updates several Together AI model YAML entries to match current provider capabilities, pricing, and lifecycle.

Qwen/Qwen3.5-397B-A17B is marked retired (status: retired, isDeprecated: true, retirementDate: 2026-06-29) with a normalized output token cost. Qwen3.7-Plus drops structured_output, tool_choice, and thinking: true, and adds removeParams: reasoning_effort. Qwen3-8B-Lora and DeepSeek-OCR-2 also strip reasoning_effort from forwarded params.

cartesia/sonic-3.5 pricing moves from zero per-token costs to input_cost_per_character. google/imagen-4.0-fast and imagen-4.0-preview now list image as an input modality and refresh documentation links.

^{Reviewed by Cursor Bugbot for commit 723bddd. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

/test-models

[harshiv-26]

Gateway test results

• Total: 11
• Passed: 2
• Failed: 0
• Validation failed: 1
• Errored: 0
• Skipped: 8
• Success rate: 66.67%

Provider	Model	Scenarios
together-ai	Qwen/Qwen3-8B-Lora	skipped: skip-check
together-ai	Qwen/Qwen3.5-397B-A17B	skipped: skip-check
together-ai	Qwen/Qwen3.7-Plus	success: params:stream, json-output:stream validation_failure: tool-call:stream skipped: tool-call, json-output, params
together-ai	cartesia/sonic-3.5	skipped: skip-check
together-ai	google/imagen-4.0-fast	skipped: skip-check
together-ai	google/imagen-4.0-preview	skipped: skip-check

Failures (1)

together-ai/Qwen/Qwen3.7-Plus — tool-call:stream (validation_failure)

Error

Traceback (most recent call last):
  File "/tmp/tmpuiv6aw4g/snippet.py", line 49, in <module>
    raise Exception("VALIDATION FAILED: tool-call stream - no tool calls received")
Exception: VALIDATION FAILED: tool-call stream - no tool calls received

Code snippet

from openai import OpenAI

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

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get the current weather for a location.",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {
                        "type": "string",
                        "description": "The city name, e.g. London",
                    },
                },
                "required": ["location"],
                "additionalProperties": False,
            },
            "strict": True,
        },
    },
]

response = client.chat.completions.create(
    model="test-v2-together-ai/Qwen-Qwen3.7-Plus",
    messages=[
        {"role": "user", "content": "Use the get_weather tool to check the weather in London. You must call the tool, do not respond with plain text."},
    ],
    tools=tools,
    tool_choice="auto",
    stream=True,
)
_tool_calls_made = False
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)
        if delta.tool_calls:
            _tool_calls_made = True
            for _tc in delta.tool_calls:
                if _tc.function:
                    print(_tc.function.arguments or "", end="", flush=True)

if not _tool_calls_made:
    raise Exception("VALIDATION FAILED: tool-call stream - no tool calls received")
print("\nVALIDATION: tool-call stream SUCCESS")

Successes (2)

together-ai/Qwen/Qwen3.7-Plus — params:stream (success)

Output

The capital of France is Paris.

together-ai/Qwen/Qwen3.7-Plus — json-output:stream (success)

Output

{"colors": [
  {"name": "Red", "hex": "#FF0000"},
  {"name": "Green", "hex": "#00FF00"},
  {"name": "Blue", "hex": "#0000FF"}
]}
VALIDATION: json-outp
... (truncated, 18 chars omitted)

Skipped (8)

together-ai/cartesia/sonic-3.5 — skip-check (skipped)

Skip reason

TTS models for Together AI are not supported

together-ai/google/imagen-4.0-fast — skip-check (skipped)

Skip reason

unsupported mode 'image'

together-ai/google/imagen-4.0-preview — skip-check (skipped)

Skip reason

unsupported mode 'image'

together-ai/Qwen/Qwen3-8B-Lora — skip-check (skipped)

Skip reason

Provisioned model

together-ai/Qwen/Qwen3.5-397B-A17B — skip-check (skipped)

Skip reason

deprecated or retired model

together-ai/Qwen/Qwen3.7-Plus — tool-call (skipped)

Skip reason

Non-stream mode is not supported by this model

together-ai/Qwen/Qwen3.7-Plus — json-output (skipped)

Skip reason

Non-stream mode is not supported by this model

together-ai/Qwen/Qwen3.7-Plus — params (skipped)

Skip reason

Non-stream mode is not supported by this model

[cursor]

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

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

^{Reviewed by Cursor Bugbot for commit 723bddd. Configure here.}