---
title: DeepSeek-V4-Flash and Qwen3.8-27B Inference LLMs Now Available on Telnyx
canonical: https://telnyx.com/release-notes/deepseek-v4-flash-qwen3-8-27b-inference
---
# DeepSeek-V4-Flash and Qwen3.8-27B Inference LLMs Now Available on Telnyx

[Release notes](https://telnyx.com/release-notes)

# DeepSeek-V4-Flash and Qwen3.8-27B Now Available for Telnyx Inference

### 21, Aug 2026

Two new smaller, faster LLMs are now available on the [Telnyx Inference API](https://telnyx.com/products/inference): DeepSeek-V4-Flash for cost-efficient high-volume workloads, and Qwen3.8-27B for low-latency routing and classification tasks. Both run on Telnyx-owned GPU infrastructure through the existing OpenAI-compatible Chat Completions endpoint, with no new endpoints or SDK updates required.

### What's new

* **DeepSeek-V4-Flash:** Available with model ID `deepseek-ai/DeepSeek-V4-Flash-0731`. A smaller, faster DeepSeek variant tuned for cost-efficient inference at high volume.
* **Qwen3.8-27B:** Available with model ID `Qwen/Qwen3.8-27B`. A 27B-parameter model from the Qwen team, sized for low-latency tasks where a full-scale LLM is overkill.
* **Drop-in integration:** Both models work with the existing OpenAI-compatible Chat Completions API. No new endpoints, no SDK updates required.

### Why it matters

* Not every agent step needs the largest available model. Routing a lightweight task to DeepSeek-V4-Flash or Qwen3.8-27B instead of a flagship LLM cuts cost per call without forcing a second inference provider into your stack.
* Running these smaller models on Telnyx-owned GPUs keeps them on the same private backbone as your STT, TTS, and telephony, so the latency hop of routing to a separate inference provider is gone.
* Evaluating open-weight models without changing your integration means zero migration cost when switching between models on the platform.

### Example use cases

* Voice AI teams routing low-complexity agent turns to a cheaper model and reserving flagship LLMs for hard reasoning steps, without a second vendor.
* Developers building classification, summarization, or intent-routing layers where a 27B model is enough and response time matters more than parameter count.
* Engineering teams running high-volume batch inference that need cost per token down without changing their existing Telnyx integration.

### Pricing

| Token Type   | DeepSeek-V4-Flash | Qwen3.8-27B       |
| ------------ | ----------------- | ----------------- |
| Input        | $0.13 / 1M tokens | $0.40 / 1M tokens |
| Cached Input | $0.03 / 1M tokens | $0.05 / 1M tokens |
| Output       | $0.26 / 1M tokens | $3.00 / 1M tokens |

Full rate details on the [inference pricing page](https://telnyx.com/pricing/inference-api).

### Getting started

1. Open the [LLM Playground](https://portal.telnyx.com/#/ai/inference/playground) in Mission Control.
2. Select `deepseek-ai/DeepSeek-V4-Flash-0731` or `Qwen/Qwen3.8-27B` from the model dropdown.
3. Send a prompt to test the model, or integrate it into your application using the [Chat Completions API](https://developers.telnyx.com/docs/inference/models).

curl https://api.telnyx.com/v2/ai/chat/completions   -H "Authorization: Bearer $TELNYX_API_KEY"   -H "Content-Type: application/json"   -d '{
    "model": "deepseek-ai/DeepSeek-V4-Flash-0731",
    "messages": [
      {"role": "user", "content": "Classify this support ticket into one of: billing, technical, account."}
    ]
  }'

**Learn more** in the [Inference API docs](https://developers.telnyx.com/docs/inference/models) or on the [pricing page](https://telnyx.com/pricing/inference-api).

[Start building](https://telnyx.com/sign-up)

### Ask AI

* [![GPT](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/chatgpt_icon-ea55e09d.jpeg?width=92&format=webp)](https://chat.openai.com/?q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.)
* [![Claude](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/claude_icon-c7c8eef3.webp?width=92&format=webp)](https://claude.ai/new?q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.)
* [![Perplexity](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/perplexity_icon-cd5b74f1.webp?width=92&format=webp)](https://www.perplexity.ai/search?q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.)
* [![Gemini](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/gemini_icon-4ff1cbf4.webp?width=92&format=webp)](https://www.google.com/search?udm=50&aep=11&q=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.)
* [![Grok](https://d2efw2akbx8gaf.cloudfront.net/https://dotcom-payload-assets-prod.s3.us-east-1.amazonaws.com/cms-media/grok_icon-e588e5bd.webp?width=92&format=webp)](https://x.com/i/grok?text=What makes Telnyx's global network different, breakdown call quality for normal phone calls and voice AI calls.)