[Bug]: API returns 404 for all model IDs and model catalog is empty

[anwarnatwankhankhamai5544-crypto]

🧩 Summary

A clear and concise description of what the bug is. Include what you expected to happen and what actually happened.

---

🔁 Steps to Reproduce

Please describe how to reproduce the problem:

If applicable, include example input, commands, or code snippets:

📸 Additional Context

Add screenshots, links, or other context about the feature request here.
Hello Bytez Team,

I need help with a Bytez API/catalog issue.

I am using a Bytez API key from a Free-plan account through a Cloudflare Worker. I have tested the API with multiple authentication formats and model IDs, but I cannot successfully access even the model shown in the API documentation.

Here is the result of my latest diagnostic test:

{
"key_length": "32",
"key_has_extra_whitespace": "false",

"run_gpt2_raw": "404: Model does not exist or has yet to be added to the Bytez catalog.",
"run_gpt2_key": "404: Model does not exist or has yet to be added to the Bytez catalog.",
"run_gpt2_bearer": "404: Model does not exist or has yet to be added to the Bytez catalog.",

"list_gpt2_raw": "200: {"error":null,"output":[]}",
"list_gpt2_key": "200: {"error":null,"output":[]}",
"list_gpt2_bearer": "200: {"error":null,"output":[]}",

"list_tasks_raw": "200: {"error":null,"output":[...]}"
}

The important point is that:

• The API key length is 32 characters.
• There is no extra whitespace in the key.
• I tested the same request with the raw API key.
• I also tested "Key <API_KEY>".
• I also tested "Bearer <API_KEY>".
• All three authentication methods produce the same 404 result.
• The model-list endpoint responds successfully with HTTP 200.
• However, querying the catalog for "openai-community/gpt2" returns an empty output array.
• The task-list endpoint works and returns the available tasks.

I have also tested several text-to-video model IDs, including:

• "openai-community/gpt2" for basic API/catalog testing
• "Wan-AI/Wan2.1-T2V-1.3B-Diffusers"
• "ali-vilab/text-to-video-ms-1.7b"
• other text-to-video model IDs

They either return the same "Model does not exist" response or cannot be found in the catalog.

What I need from Bytez

Could you please confirm the following:

1. Is the Bytez model catalog currently working correctly?
2. Is the "/models/v2/{modelId}" API endpoint still the correct endpoint?
3. Is the authentication header format still the same?
4. What is the exact model ID of one currently working text-to-video model that can be called through the API?
5. Is that model available to a Free-plan account?
6. What is the exact request format for that model?
7. If the model requires a different endpoint or API version, please provide the current endpoint.
8. If text-to-video is not currently available through the Free plan, please state that clearly.
9. If the catalog is currently being migrated or the API has changed, please tell me what the current integration method is.

My actual goal

I am developing a small AI video-generation project called Xindun VideoForge AI.

The intended architecture is:

User prompt
→ Cloudflare Worker
→ Bytez API
→ open-source text-to-video model
→ generated video/result
→ Xindun VideoForge AI

I am specifically looking for a currently working open-source text-to-video model/API route that I can use without automatic paid charges.

I do NOT need a list of theoretical or old model IDs.

I need one currently working, verified example that I can test from my Cloudflare Worker.

If possible, please provide a complete working example containing:

• exact model ID
• exact API endpoint
• exact authentication header
• exact JSON request body
• expected response
• whether generation is synchronous or asynchronous
• polling endpoint if applicable
• how to retrieve the generated video
• current Free-plan limits/credits

I have already spent significant time testing different model IDs and authentication formats, so I would greatly appreciate a definitive current answer.

Thank you for your help.

Best regards,
Anwar Khan
Xindun VideoForge AI

[anwarnatwankhankhamai5544-crypto]

Hello Bytez Team,

I need help with a Bytez API/catalog issue.

I am using a Bytez API key from a Free-plan account through a Cloudflare Worker. I have tested the API with multiple authentication formats and model IDs, but I cannot successfully access even the model shown in the API documentation.

Here is the result of my latest diagnostic test:

{
"key_length": "32",
"key_has_extra_whitespace": "false",

"run_gpt2_raw": "404: Model does not exist or has yet to be added to the Bytez catalog.",
"run_gpt2_key": "404: Model does not exist or has yet to be added to the Bytez catalog.",
"run_gpt2_bearer": "404: Model does not exist or has yet to be added to the Bytez catalog.",

"list_gpt2_raw": "200: {"error":null,"output":[]}",
"list_gpt2_key": "200: {"error":null,"output":[]}",
"list_gpt2_bearer": "200: {"error":null,"output":[]}",

"list_tasks_raw": "200: {"error":null,"output":[...]}"
}

The important point is that:

• The API key length is 32 characters.
• There is no extra whitespace in the key.
• I tested the same request with the raw API key.
• I also tested "Key <API_KEY>".
• I also tested "Bearer <API_KEY>".
• All three authentication methods produce the same 404 result.
• The model-list endpoint responds successfully with HTTP 200.
• However, querying the catalog for "openai-community/gpt2" returns an empty output array.
• The task-list endpoint works and returns the available tasks.

I have also tested several text-to-video model IDs, including:

• "openai-community/gpt2" for basic API/catalog testing
• "Wan-AI/Wan2.1-T2V-1.3B-Diffusers"
• "ali-vilab/text-to-video-ms-1.7b"
• other text-to-video model IDs

They either return the same "Model does not exist" response or cannot be found in the catalog.

What I need from Bytez

Could you please confirm the following:

1. Is the Bytez model catalog currently working correctly?
2. Is the "/models/v2/{modelId}" API endpoint still the correct endpoint?
3. Is the authentication header format still the same?
4. What is the exact model ID of one currently working text-to-video model that can be called through the API?
5. Is that model available to a Free-plan account?
6. What is the exact request format for that model?
7. If the model requires a different endpoint or API version, please provide the current endpoint.
8. If text-to-video is not currently available through the Free plan, please state that clearly.
9. If the catalog is currently being migrated or the API has changed, please tell me what the current integration method is.

My actual goal

I am developing a small AI video-generation project called Xindun VideoForge AI.

The intended architecture is:

User prompt
→ Cloudflare Worker
→ Bytez API
→ open-source text-to-video model
→ generated video/result
→ Xindun VideoForge AI

I am specifically looking for a currently working open-source text-to-video model/API route that I can use without automatic paid charges.

I do NOT need a list of theoretical or old model IDs.

I need one currently working, verified example that I can test from my Cloudflare Worker.

If possible, please provide a complete working example containing:

• exact model ID
• exact API endpoint
• exact authentication header
• exact JSON request body
• expected response
• whether generation is synchronous or asynchronous
• polling endpoint if applicable
• how to retrieve the generated video
• current Free-plan limits/credits

I have already spent significant time testing different model IDs and authentication formats, so I would greatly appreciate a definitive current answer.

Thank you for your help.

Best regards,
Anwar Khan
Xindun VideoForge AI

[anwarnatwankhankhamai5544-crypto]

Hello Bytez Team,

I need help with a Bytez API/catalog issue.

I am using a Bytez API key from a Free-plan account through a Cloudflare Worker. I have tested the API with multiple authentication formats and model IDs, but I cannot successfully access even the model shown in the API documentation.

Here is the result of my latest diagnostic test:

{
"key_length": "32",
"key_has_extra_whitespace": "false",

"run_gpt2_raw": "404: Model does not exist or has yet to be added to the Bytez catalog.",
"run_gpt2_key": "404: Model does not exist or has yet to be added to the Bytez catalog.",
"run_gpt2_bearer": "404: Model does not exist or has yet to be added to the Bytez catalog.",

"list_gpt2_raw": "200: {"error":null,"output":[]}",
"list_gpt2_key": "200: {"error":null,"output":[]}",
"list_gpt2_bearer": "200: {"error":null,"output":[]}",

"list_tasks_raw": "200: {"error":null,"output":[...]}"
}

The important point is that:

• The API key length is 32 characters.
• There is no extra whitespace in the key.
• I tested the same request with the raw API key.
• I also tested "Key <API_KEY>".
• I also tested "Bearer <API_KEY>".
• All three authentication methods produce the same 404 result.
• The model-list endpoint responds successfully with HTTP 200.
• However, querying the catalog for "openai-community/gpt2" returns an empty output array.
• The task-list endpoint works and returns the available tasks.

I have also tested several text-to-video model IDs, including:

• "openai-community/gpt2" for basic API/catalog testing
• "Wan-AI/Wan2.1-T2V-1.3B-Diffusers"
• "ali-vilab/text-to-video-ms-1.7b"
• other text-to-video model IDs

They either return the same "Model does not exist" response or cannot be found in the catalog.

What I need from Bytez

Could you please confirm the following:

1. Is the Bytez model catalog currently working correctly?
2. Is the "/models/v2/{modelId}" API endpoint still the correct endpoint?
3. Is the authentication header format still the same?
4. What is the exact model ID of one currently working text-to-video model that can be called through the API?
5. Is that model available to a Free-plan account?
6. What is the exact request format for that model?
7. If the model requires a different endpoint or API version, please provide the current endpoint.
8. If text-to-video is not currently available through the Free plan, please state that clearly.
9. If the catalog is currently being migrated or the API has changed, please tell me what the current integration method is.

My actual goal

I am developing a small AI video-generation project called Xindun VideoForge AI.

The intended architecture is:

User prompt
→ Cloudflare Worker
→ Bytez API
→ open-source text-to-video model
→ generated video/result
→ Xindun VideoForge AI

I am specifically looking for a currently working open-source text-to-video model/API route that I can use without automatic paid charges.

I do NOT need a list of theoretical or old model IDs.

I need one currently working, verified example that I can test from my Cloudflare Worker.

If possible, please provide a complete working example containing:

• exact model ID
• exact API endpoint
• exact authentication header
• exact JSON request body
• expected response
• whether generation is synchronous or asynchronous
• polling endpoint if applicable
• how to retrieve the generated video
• current Free-plan limits/credits

I have already spent significant time testing different model IDs and authentication formats, so I would greatly appreciate a definitive current answer.

Thank you for your help.

Best regards,
Anwar Khan
Xindun VideoForge AI