New Models Added: Eleven v3, Kimi k2.5, and Qwen3 Updates

Date: 2026-02-05 / (1404-11-16)

Summary

We have added support for ElevenLabs' latest Eleven v3 model, Moonshot AI's Kimi k2.5, and Alibaba's Qwen3-Rerank model. Additionally, we have updated the Qwen3-Max and Qwen3-VL-Flash models with their latest snapshots, offering improved performance at the same price.

---

Details

ElevenLabs

Eleven v3

Eleven v3 (eleven_v3) is ElevenLabs' most advanced speech synthesis model, delivering emotionally rich and lifelike speech. It supports natural multi-speaker dialogue and offers a high degree of contextual understanding.

• Endpoint: v1/audio/speech
• Capabilities: High emotional range, multi-speaker dialogue support, 70+ languages.
• Pricing: $0.005 per second of generated audio.

Pricing Details:

Model	Input	Cached Input	Output	Pricing Unit
eleven_v3	-	-	-	$0.005 / second

Example Request

curl https://api.avalai.ir/v1/audio/speech \
  -H "Authorization: Bearer $AVALAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "eleven_v3",
    "input": "This is a demonstration of the new Eleven v3 model.",
    "voice": "coral"
  }' \
  --output test_audio.mp3

Moonshot AI

Kimi k2.5

Kimi k2.5 (kimi-k2.5) is a powerful multimodal model from Moonshot AI, designed for advanced reasoning, coding, and agentic tasks. It excels in complex workflows and supports vision capabilities.

• Endpoints: v1/chat/completions, v1/responses
• Capabilities: Advanced reasoning, coding, vision, agentic behavior.

Pricing Details:

Model	Input	Cached Input	Output
kimi-k2.5	$0.60	$0.10	$3.00
(Prices per 1M tokens)

Example Request

curl https://api.avalai.ir/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AVALAI_API_KEY" \
  -d '{
    "model": "kimi-k2.5",
    "messages": [
      {
        "role": "user",
        "content": "Explain the significance of the Kimi k2.5 model."
      }
    ]
  }'

Example Response

{
  "id": "chatcmpl-kimi-example",
  "created": 1709673400,
  "model": "kimi-k2.5",
  "object": "chat.completion",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Kimi k2.5 is a significant advancement in..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 25,
    "total_tokens": 37
  }
}

Alibaba

Qwen3 Rerank

Qwen3 Rerank (qwen3-rerank) is a specialized model for re-ranking text documents, essential for RAG (Retrieval-Augmented Generation) applications to improve retrieval accuracy.

• Endpoint: v1/rerank
• Capabilities: High-accuracy text reranking, supports 100+ languages.

Pricing Details:

Model	Input	Cached Input	Output
qwen3-rerank	$0.10	$0.0035	-
(Prices per 1M tokens)

Example Request

curl https://api.avalai.ir/v1/rerank \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AVALAI_API_KEY" \
  -d '{
    "model": "qwen3-rerank",
    "query": "What is machine learning?",
    "documents": [
      "Machine learning is a field of study...",
      "Deep learning is a subset of machine learning...",
      "Apples are a type of fruit..."
    ]
  }'

Example Response

{
  "model": "qwen3-rerank",
  "results": [
    {
      "index": 1,
      "relevance_score": 0.98,
      "document": {
        "text": "Deep learning is a subset of machine learning..."
      }
    },
    {
      "index": 0,
      "relevance_score": 0.95,
      "document": {
        "text": "Machine learning is a field of study..."
      }
    },
    {
      "index": 2,
      "relevance_score": 0.01,
      "document": {
        "text": "Apples are a type of fruit..."
      }
    }
  ],
  "usage": {
    "total_tokens": 45
  }
}

Qwen3 Model Updates

We have also updated the following Qwen3 models to their latest snapshots. These updates provide improved performance while maintaining the same pricing.

• Qwen3 Max: Updated to qwen3-max-2026-01-23
• Qwen3 VL Flash: Updated to qwen3-vl-flash-2026-01-22

You can use these specific snapshot names to pin your application to these versions, or continue using the base model names (qwen3-max, qwen3-vl-flash) to automatically use the latest stable versions.

Pricing Details (Same as previous versions):

Model	Input	Cached Input	Output	Special Pricing
qwen3-max-2026-01-23	$1.20	$0.10	$6.00	Higher rates for >32K and >128K context
qwen3-vl-flash-2026-01-22	$0.05	$0.01	$0.40	Higher rates for >32K and >128K context

(Prices per 1M tokens. See pricing page for detailed context tiers.)