# No sufficient capacity to run baai\_bge\_m3 model

**URL:** <https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693>\
**Category:** Elasticsearch\
**Tags:** elastic-stack-machine-learning\
**Created:** [October 14, 2025, 10:02am UTC](https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693 "2025-10-14T10:02:10Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![gueri](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/gueri/32/103243_2.png) [@gueri](https://discuss.elastic.co/u/gueri)\
**Post date:** [October 14, 2025, 10:02am UTC](https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693/1 "2025-10-14T10:02:10Z")

</div>

Hi,

My ML configuration is 2 nodes with this configuration for each node :

> **ml.allocated\_processors\_double** 8.0
> 
> **ml.machine\_memory** 62.4GB
> 
> **ml.config\_version** 12.0.0
> 
> **ml.max\_jvm\_size** 4GB
> 
> **ml.allocated\_processors** 8

I already use model bge-large-en-v1.5 for our RAG but now I want to test baai\_bge\_m3 [BGE-M3 — BGE documentation](https://bge-model.com/bge/bge_m3.html#) because this model can deal with more 512 tokens.

The upload of the model baai\_bge\_m3 is OK.

I stopped all runnings models so all resources are free and I tried to start the model baai\_bge\_m3. But I have this issue.

```auto
{
  "error": {
    "root_cause": [
      {
        "type": "status_exception",
        "reason": "Could not start deployment because no ML nodes with sufficient capacity were found"
      }
    ],
    "type": "status_exception",
    "reason": "Could not start deployment because no ML nodes with sufficient capacity were found",
    "caused_by": {
      "type": "illegal_state_exception",
      "reason": "Could not start deployment because no suitable nodes were found, allocation explanation [none]"
    }
  },
  "status": 429
}

```

I suspect “sufficient capacity” means no enough memory ???

When I use bge-large-en-v1 with 4 allocations/1 thread “Model size stats” is :

model\_size\_bytes 1.2GB  
required\_native\_memory\_bytes 10.2GB

BGE-M3 documentation says model size is 2,27 GB but I have no idea how much memory is needed to run this model on ELK.

I appreciate any help 🙂

Regards,

---

<div class="post-metadata">

**Author:** ![gueri](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/gueri/32/103243_2.png) [@gueri](https://discuss.elastic.co/u/gueri)\
**Post date:** [October 14, 2025, 1:19pm UTC](https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693/2 "2025-10-14T13:19:27Z")

</div>

I probably found a clue.

It seems that the memory size used by ML node is limited by default to 30% of the available memory. I will ask our Elastic team to check this parameter in our cluster…

---

<div class="post-metadata">

**Author:** ![dkyle](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/dkyle/32/59114_2.png) [@dkyle](https://discuss.elastic.co/u/dkyle)\
**Post date:** [October 14, 2025, 2:24pm UTC](https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693/3 "2025-10-14T14:24:28Z")

</div>

Most likely it is due to the [ML settings](https://www.elastic.co/docs/reference/elasticsearch/configuration-reference/machine-learning-settings).

Try setting `xpack.ml.use_auto_machine_memory_percent: true`

---

<div class="post-metadata">

**Author:** ![gueri](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/gueri/32/103243_2.png) [@gueri](https://discuss.elastic.co/u/gueri)\
**Post date:** [October 14, 2025, 3:19pm UTC](https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693/4 "2025-10-14T15:19:39Z")

</div>

Yes David, you’re right, it is probably this setting.

I’m waiting response from our elastic team that manage the cluster.

---

<div class="post-metadata">

**Author:** ![gueri](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/gueri/32/103243_2.png) [@gueri](https://discuss.elastic.co/u/gueri)\
**Post date:** [November 3, 2025, 1:37pm UTC](https://discuss.elastic.co/t/no-sufficient-capacity-to-run-baai-bge-m3-model/382693/5 "2025-11-03T13:37:30Z")

</div>

There you go, it’s done! The option has been changed on our cluster and I can now load the model. However, I see that this will raise another problem!

Our servers have 64 GB of memory. With a single allocation and a single thread, the model consumes practically all the memory of a ML node. In production, I can’t be satisfied with a single allocation. The memory size of my ML nodes will be far insufficient! I will need a less demanding model with at least 2,000 tokens… but does such an embeddings model exist ?