Incorrect response from "anthropic/claude-3.5-sonnet" model on OpenRouter API

[ghominejad]

Describe the bug
When using the OpenAI format API request with the "anthropic/claude-3.5-sonnet" model via the endpoint https://openrouter.ai/api/v1/chat/completions, the response completion is not generating properly.

To Reproduce
Steps to reproduce the behavior:

1. Make an API request to the endpoint https://openrouter.ai/api/v1/chat/completions using the OpenAI standard format.
2. Use the model "anthropic/claude-3.5-sonnet".
3. Example prompt:
   • System: You are Claude, a large language model trained by Anthropic. Carefully heed the user's instructions. Respond very concise.
   • User: loop from 1 to 100 in python
4. Observe the response.

Expected behavior
The response should generate a concise and correct Python loop for iterating from 1 to 100.

Actual behavior
The response is incorrect and incomplete. Example response:

Here's a concise Python loop

for1, 101(i)

Additional context
This issue seems to be specific to the "anthropic/claude-3.5-sonnet" model when invoked using the OpenAI standard format. Other models or formats might not exhibit this behavior.

---

You can submit this report on the OpenRouter GitHub issues page.