github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 08:02:14 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #6462] Make tool call response compatible with OpenAI format #66102

New Issue
Closed
opened 2026-05-03 23:58:39 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-05-03 23:58:39 -05:00
Owner
Copy Link

Originally created by @eliasfroehner on GitHub (Aug 22, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6462

Description

Currently, the response for a tool call looks like this:

{
  "model": "llama3.1",
  "created_at": "2024-07-22T20:33:28.123648Z",
  "message": {
    "role": "assistant",
    "content": "",
    "tool_calls": [
      {
        "function": {
          "name": "get_current_weather",
          "arguments": {
            "format": "celsius",
            "location": "Paris, FR"
          }
        }
      }
    ]
  },
  "done_reason": "stop",
  "done": true,
  "total_duration": 885095291,
  "load_duration": 3753500,
  "prompt_eval_count": 122,
  "prompt_eval_duration": 328493000,
  "eval_count": 33,
  "eval_duration": 552222000
}

However, I would like the response to be compatible with the OpenAI format, which includes an id and type field for each tool call. The desired format is as follows:

{
  "model": "llama3.1",
  "created_at": "2024-07-22T20:33:28.123648Z",
  "message": {
    "role": "assistant",
    "content": "",
    "tool_calls": [
      {
        "id": "call_62136354",
        "type":  "function",
        "function": {
          "name": "get_current_weather",
          "arguments": {
            "format": "celsius",
            "location": "Paris, FR"
          }
        }
      }
    ]
  },
  "done_reason": "stop",
  "done": true,
  "total_duration": 885095291,
  "load_duration": 3753500,
  "prompt_eval_count": 122,
  "prompt_eval_duration": 328493000,
  "eval_count": 33,
  "eval_duration": 552222000
}

Additionally, the lack of an id field in the current format prevents a software from responding to multiple simultaneous calls to a tool.

Steps to Reproduce

  1. Make a tool call request.
  2. Observe the response format.

Expected Behavior

The response should include an id and type field for each tool call, making it compatible with the OpenAI format.

Actual Behavior

The response does not include an id and type field for each tool call.

Additional Information

This change is necessary to ensure compatibility with the OpenAI API and to handle multiple simultaneous tool calls effectively.

Labels

  • enhancement
  • compatibility
  • api
Originally created by @eliasfroehner on GitHub (Aug 22, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6462 ### Description Currently, the response for a tool call looks like this: ```json { "model": "llama3.1", "created_at": "2024-07-22T20:33:28.123648Z", "message": { "role": "assistant", "content": "", "tool_calls": [ { "function": { "name": "get_current_weather", "arguments": { "format": "celsius", "location": "Paris, FR" } } } ] }, "done_reason": "stop", "done": true, "total_duration": 885095291, "load_duration": 3753500, "prompt_eval_count": 122, "prompt_eval_duration": 328493000, "eval_count": 33, "eval_duration": 552222000 } ``` However, I would like the response to be compatible with the OpenAI format, which includes an `id` and `type` field for each tool call. The desired format is as follows: ```json { "model": "llama3.1", "created_at": "2024-07-22T20:33:28.123648Z", "message": { "role": "assistant", "content": "", "tool_calls": [ { "id": "call_62136354", "type": "function", "function": { "name": "get_current_weather", "arguments": { "format": "celsius", "location": "Paris, FR" } } } ] }, "done_reason": "stop", "done": true, "total_duration": 885095291, "load_duration": 3753500, "prompt_eval_count": 122, "prompt_eval_duration": 328493000, "eval_count": 33, "eval_duration": 552222000 } ``` Additionally, the lack of an `id` field in the current format prevents a software from responding to multiple simultaneous calls to a tool. ### Steps to Reproduce 1. Make a tool call request. 2. Observe the response format. ### Expected Behavior The response should include an `id` and `type` field for each tool call, making it compatible with the OpenAI format. ### Actual Behavior The response does not include an `id` and `type` field for each tool call. ### Additional Information This change is necessary to ensure compatibility with the OpenAI API and to handle multiple simultaneous tool calls effectively. ### Labels - enhancement - compatibility - api
GiteaMirror added the feature request label 2026-05-03 23:58:40 -05:00
GiteaMirror commented 2026-05-03 23:58:44 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 22, 2024):

If you use the openai compatibility endpoint (/v1/chat/completions), you get a response with id and type fields. Don't know how compatible the rest of the structure is, eg the arguments are a string rather than a dict.

{
  "id": "chatcmpl-625",
  "object": "chat.completion",
  "created": 1724330640,
  "model": "llama3.1",
  "system_fingerprint": "fp_ollama",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "",
        "tool_calls": [
          {
            "id": "call_5dytv9km",
            "type": "function",
            "function": {
              "name": "get_current_weather",
              "arguments": "{\"format\":\"celsius\",\"location\":\"Paris\"}"
            }
          }
        ]
      },
      "finish_reason": "tool_calls"
    }
  ],
  "usage": {
    "prompt_tokens": 204,
    "completion_tokens": 25,
    "total_tokens": 229
  }
}
<!-- gh-comment-id:2304598853 --> @rick-github commented on GitHub (Aug 22, 2024): If you use the openai compatibility endpoint (`/v1/chat/completions`), you get a response with `id` and `type` fields. Don't know how compatible the rest of the structure is, eg the `arguments` are a string rather than a dict. ```json { "id": "chatcmpl-625", "object": "chat.completion", "created": 1724330640, "model": "llama3.1", "system_fingerprint": "fp_ollama", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "", "tool_calls": [ { "id": "call_5dytv9km", "type": "function", "function": { "name": "get_current_weather", "arguments": "{\"format\":\"celsius\",\"location\":\"Paris\"}" } } ] }, "finish_reason": "tool_calls" } ], "usage": { "prompt_tokens": 204, "completion_tokens": 25, "total_tokens": 229 } } ```
GiteaMirror commented 2026-05-03 23:58:46 -05:00
Author
Owner
Copy Link

@eliasfroehner commented on GitHub (Aug 22, 2024):

@rick-github Thank you very much for pointing out.

<!-- gh-comment-id:2304757159 --> @eliasfroehner commented on GitHub (Aug 22, 2024): @rick-github Thank you very much for pointing out.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#66102
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?