Ollama model list and aider model list differ

[ahonnecke]

Issue

Aider fails to fetch api/show.

However, CURL seems to do so just fine.

ahonnecke@antonym:~/src/jolly-brancher$ curl http://10.0.1.111:11434/api/show -d '{
  "model": "qwen2.5-coder:32b"
}'
{"license":"\n                                 Apache License\n                           Version 2.0, January 2004\n                        http://www.apache.org/licenses/\n\n   TERMS 

The model list that ollama delivers seems to not match the list that aider delivers.

ahonnecke@antonym:~/src/jolly-brancher$ aider --list-models ollama/
──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Models which match "ollama/":
- ollama/codegeex4
- ollama/deepseek-coder-v2-instruct
- ollama/deepseek-coder-v2-lite-instruct
- ollama/internlm2_5-20b-chat
- ollama/llama2
- ollama/llama2:13b
- ollama/llama2:70b
- ollama/llama2:7b
- ollama/llama3
- ollama/llama3.1
- ollama/llama3:70b
- ollama/llama3:8b
- ollama/mistral-7B-Instruct-v0.1
- ollama/mistral-7B-Instruct-v0.2
- ollama/mistral-large-instruct-2407
- ollama/mixtral-8x22B-Instruct-v0.1
- ollama/mixtral-8x7B-Instruct-v0.1

ahonnecke@antonym:~/src/jolly-brancher$ curl http://10.0.1.111:11434/api/tags | jq
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  1015  100  1015    0     0   159k      0 --:--:-- --:--:-- --:--:--  165k
{
  "models": [
    {
      "name": "qwen2.5-coder:32b",
      "model": "qwen2.5-coder:32b",
      "modified_at": "2025-01-29T09:25:49.572611118-07:00",
      "size": 19851349856,
      "digest": "4bd6cbf2d094264457a17aab6bd6acd1ed7a72fb8f8be3cfb193f63c78dd56df",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "qwen2",
        "families": [
          "qwen2"
        ],
        "parameter_size": "32.8B",
        "quantization_level": "Q4_K_M"
      }
    },
    {
      "name": "qwen:32b",
      "model": "qwen:32b",
      "modified_at": "2025-01-29T08:07:07.888901163-07:00",
      "size": 18498677860,
      "digest": "26e7e8447f5d7fba43b4bc11236fc6f9db4e19ff3184f305b39c7ca76eb896a1",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "qwen2",
        "families": [
          "qwen2"
        ],
        "parameter_size": "33B",
        "quantization_level": "Q4_0"
      }
    },
    {
      "name": "deepseek-r1:32b",
      "model": "deepseek-r1:32b",
      "modified_at": "2025-01-28T14:40:59.484812837-07:00",
      "size": 19851337640,
      "digest": "38056bbcbb2d068501ecb2d5ea9cea9dd4847465f1ab88c4d4a412a9f7792717",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "qwen2",
        "families": [
          "qwen2"
        ],
        "parameter_size": "32.8B",
        "quantization_level": "Q4_K_M"
      }
    }
  ]
}

[spaasis]

The /api/show is a POST endpoint, which might be why you get 404 on the local curl run with GET.
https://github.com/ollama/ollama/blob/main/docs/api.md#show-model-information

No idea why Aider fails here though, just a heads up

[ahonnecke]

Thank you @spaasis!

the POST does indeed work.

  "model": "qwen2.5-coder:32b"
}'
{"license":"\n                                 Apache License\n                           Version 2.0, January 2004\n                        http://www.apache.org/licenses/\n\n   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION\n\n   1. Definitions.\n\n      \"License\" shall mean the terms and conditions for use, reproduction,\n      and distribution as defined by Sections 1 through 9 of this document.\n\n      \"Licensor\" shall mean the copyright owner or entity authorized by\n      the copyright owner that is granting the License.\n\n      \"Legal Entity\" shall mean the union of the acting entity and all\n      other entities that control, are controlled by, or are under common\n      control with that entity. For the purposes of this definition,\n      \"control\" means (i) the power, direct or indirect, to cause the\n      direction or management of such entity, whether by contract or\n      otherwise, or (ii) ownership of fifty percent (50%) or more of the\n      outstanding shares, or (iii) beneficial ownership of such entity.\n\n      \"You\" (or \"Your\") shall mean an individual or Legal Entity\n      exercising permissions granted by this License.\n\n      \"Source\" form shall mean the preferred form for making modifications,\n      including but not limited to software source code, documentation\n      source, and configuration files.\n\n```

[ahonnecke]

The model list that ollama delivers seems to not match the list that aider delivers.

ahonnecke@antonym:~/src/jolly-brancher$ aider --list-models ollama/
──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Models which match "ollama/":
- ollama/codegeex4
- ollama/deepseek-coder-v2-instruct
- ollama/deepseek-coder-v2-lite-instruct
- ollama/internlm2_5-20b-chat
- ollama/llama2
- ollama/llama2:13b
- ollama/llama2:70b
- ollama/llama2:7b
- ollama/llama3
- ollama/llama3.1
- ollama/llama3:70b
- ollama/llama3:8b
- ollama/mistral-7B-Instruct-v0.1
- ollama/mistral-7B-Instruct-v0.2
- ollama/mistral-large-instruct-2407
- ollama/mixtral-8x22B-Instruct-v0.1
- ollama/mixtral-8x7B-Instruct-v0.1

ahonnecke@antonym:~/src/jolly-brancher$ curl http://10.0.1.111:11434/api/tags | jq
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  1015  100  1015    0     0   159k      0 --:--:-- --:--:-- --:--:--  165k
{
  "models": [
    {
      "name": "qwen2.5-coder:32b",
      "model": "qwen2.5-coder:32b",
      "modified_at": "2025-01-29T09:25:49.572611118-07:00",
      "size": 19851349856,
      "digest": "4bd6cbf2d094264457a17aab6bd6acd1ed7a72fb8f8be3cfb193f63c78dd56df",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "qwen2",
        "families": [
          "qwen2"
        ],
        "parameter_size": "32.8B",
        "quantization_level": "Q4_K_M"
      }
    },
    {
      "name": "qwen:32b",
      "model": "qwen:32b",
      "modified_at": "2025-01-29T08:07:07.888901163-07:00",
      "size": 18498677860,
      "digest": "26e7e8447f5d7fba43b4bc11236fc6f9db4e19ff3184f305b39c7ca76eb896a1",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "qwen2",
        "families": [
          "qwen2"
        ],
        "parameter_size": "33B",
        "quantization_level": "Q4_0"
      }
    },
    {
      "name": "deepseek-r1:32b",
      "model": "deepseek-r1:32b",
      "modified_at": "2025-01-28T14:40:59.484812837-07:00",
      "size": 19851337640,
      "digest": "38056bbcbb2d068501ecb2d5ea9cea9dd4847465f1ab88c4d4a412a9f7792717",
      "details": {
        "parent_model": "",
        "format": "gguf",
        "family": "qwen2",
        "families": [
          "qwen2"
        ],
        "parameter_size": "32.8B",
        "quantization_level": "Q4_K_M"
      }
    }
  ]
}

[ahonnecke]

Looks like aider returns models that I don't have.

I can run models that exist in aider and that I have running locally:

Aider behaves the same way if I try to use a model that:

• Does not exist

• Exists, but is not on my server

So... i think that maybe Aider just needs support for qwen2.5-coder:32b added?

[ahonnecke]

I seem to have a version that's more recent than the most recent release on gihub?

aider 0.72.3

[ahonnecke]

Aider client:

LLM Host:

Aider will start and use --model ollama_chat/deepseek-r1:32b, but not ollama_chat/qwen2.5-coder

[dkhokhlov]

I see the same issue. it seems that aider is listing known internal models, that have nothing to do with ollama list.
and aider works if I start it with real model name that I have:

--model ollama/<real model name in ollam>

if you inject a typo into the name - aider will complain.