# LLM Model access

**URL:** <https://community.sambanova.ai/t/llm-model-access/1362>\
**Category:** Discussion\
**Tags:** dev, support-knowledge\
**Created:** [September 15, 2025, 10:28am UTC](https://community.sambanova.ai/t/llm-model-access/1362 "2025-09-15T10:28:45Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![bharatkumar](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/bharatkumar/32/2871_2.png) [@bharatkumar](https://community.sambanova.ai/u/bharatkumar)\
**Post date:** [September 15, 2025, 10:28am UTC](https://community.sambanova.ai/t/llm-model-access/1362/1 "2025-09-15T10:28:45Z")

</div>

Can we get access to qwen2.5-vl-72b model on request?

---

<div class="post-metadata">

**Author:** ![prajwal.balapure](https://avatars.discourse-cdn.com/v4/letter/p/a9a28c/32.png) [@prajwal.balapure](https://community.sambanova.ai/u/prajwal.balapure)\
**Post date:** [September 16, 2025, 8:31am UTC](https://community.sambanova.ai/t/llm-model-access/1362/2 "2025-09-16T08:31:08Z")

</div>

Hello Bharat,  
Thanks for reaching out! We’ve passed your request along to our internal team and will get back to you once we have an update.

---

<div class="post-metadata">

**Author:** ![prajwal.balapure](https://avatars.discourse-cdn.com/v4/letter/p/a9a28c/32.png) [@prajwal.balapure](https://community.sambanova.ai/u/prajwal.balapure)\
**Post date:** [September 19, 2025, 1:56pm UTC](https://community.sambanova.ai/t/llm-model-access/1362/3 "2025-09-19T13:56:49Z")

</div>

Hello Bharat,

Thanks for your request regarding the `qwen2.5-vl-72b` **model**. This model has been **deprecated** , but we recommend the following alternatives as potential replacements (see the last section of our [Model deprecations - SambaNova Documentation](https://docs.sambanova.ai/docs/en/models/deprecations)): -

- **Meta-Llama-3.3-70B-Instruct**
- **DeepSeek-V3-0324**

If you still specifically need access to `qwen2.5-vl-72b`, we are engaging our sales team to check the feasibility and will update you as soon as we have more information.

---

<div class="post-metadata">

**Author:** ![bharatkumar](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/bharatkumar/32/2871_2.png) [@bharatkumar](https://community.sambanova.ai/u/bharatkumar)\
**Post date:** [September 22, 2025, 6:28am UTC](https://community.sambanova.ai/t/llm-model-access/1362/4 "2025-09-22T06:28:41Z")

</div>

Thanks for the update, we are interested in Vision models and we found Qwen2.5-vl-72b works best on scanned images.

---

<div class="post-metadata">

**Author:** ![prajwal.balapure](https://avatars.discourse-cdn.com/v4/letter/p/a9a28c/32.png) [@prajwal.balapure](https://community.sambanova.ai/u/prajwal.balapure)\
**Post date:** [September 25, 2025, 7:21pm UTC](https://community.sambanova.ai/t/llm-model-access/1362/5 "2025-09-25T19:21:30Z")

</div>

Sure, we have already engaged the sales team for this.  
Thank you.