What model collections does XALEN offer?

XALEN offers 7 curated collections: Faith-Tech Specialist (Vedika models for astrology/temples), Fastest Inference (sub-200ms models), Code Generation, Image Creation, Voice & Audio, Embeddings & Search, and Budget-Friendly. Each collection is hand-picked for specific workloads.

Can I use models from different collections together?

Yes. All 200+ models are available through the same API. Collections are just curated groupings to help you discover the right model. You can mix models from any collection in your application. Just change the model parameter in your API call.

Which collection should I use for temple apps?

Start with the Faith-Tech Specialist collection. It includes Vedika Standard for complex astrology, Vedika Standard for everyday queries, Vedika Fast for real-time chat, and voice models for multilingual devotee interaction. Add models from other collections as needed.

Model Collections

Best for Faith-Tech

Models tuned for religious content, astrology computation, devotional text, and temple operations. Includes XALEN's proprietary Vedika models alongside top general-purpose LLMs.

8

Models

$0.28

Avg / 1M tok

Vedika Standard XALEN $0.60 / 1M input

Vedika Standard XALEN $2.00 / 1M input

Llama 4 Maverick Meta $0.20 / 1M input

Qwen 3 235B Alibaba $0.20 / 1M input

DeepSeek R1 DeepSeek $0.17 / 1M input

Gemma 3 27B Google $0.10 / 1M input

Mistral Large 2 Mistral $0.20 / 1M input

Vedika Swift XALEN $0.10 / 1M input

Fastest Inference

Optimized for latency-critical applications: real-time chat, voice assistants, live dashboards. Sub-100ms time-to-first-token on most models.

7

Models

$0.07

Avg / 1M tok

Vedika Swift XALEN $0.10 / 1M input

Llama 4 Scout Meta $0.08 / 1M input

Gemma 3 12B Google $0.05 / 1M input

Qwen 3 32B Alibaba $0.07 / 1M input

Mistral Small 3.1 Mistral $0.04 / 1M input

Llama 3.3 70B Meta $0.06 / 1M input

Phi-4 14B Microsoft $0.04 / 1M input

Best for Code

Code generation, completion, review, and debugging. Top performers on HumanEval, SWE-bench, and real-world coding tasks.

6

Models

$0.23

Avg / 1M tok

DeepSeek R1 DeepSeek $0.17 / 1M input

Qwen 3 235B Alibaba $0.20 / 1M input

Llama 4 Maverick Meta $0.20 / 1M input

Mistral Large 2 Mistral $0.20 / 1M input

Qwen 2.5 Coder 32B Alibaba $0.07 / 1M input

DeepSeek V3 DeepSeek $0.13 / 1M input

Image Generation

Text-to-image, image-to-image, inpainting, and upscaling. From photorealistic to artistic styles.

5

Models

$0.03

Avg / image

FLUX 1.1 Pro Ultra Black Forest Labs $0.06 / image

FLUX 1.1 Pro Black Forest Labs $0.04 / image

Stable Diffusion 3.5 Large Stability AI $0.04 / image

FLUX Schnell Black Forest Labs $0.003 / image

Stable Diffusion XL Stability AI $0.002 / image

Voice & Audio

Text-to-speech, speech-to-text, and voice cloning. 31 languages including 14 Indian languages with native speaker quality.

5

Models

$0.01

Avg / min

Vedika Voice (TTS) XALEN $0.015 / 1K chars

Vedika Voice (STT) XALEN $0.006 / minute

Whisper Large V3 OpenAI $0.006 / minute

TTS-1 HD Multi-provider $0.030 / 1K chars

TTS-1 Multi-provider $0.015 / 1K chars

Embeddings

Vector embeddings for search, RAG, clustering, and classification. Multilingual support with up to 8192 token context.

4

Models

$0.01

Avg / 1M tok

E5 Large V2 Multi-provider $0.008 / 1M tokens

BGE Large EN V1.5 BAAI $0.008 / 1M tokens

GTE Large Alibaba $0.008 / 1M tokens

UAE Large V1 WhereIsAI $0.008 / 1M tokens

Budget-Friendly

Maximum capability per dollar. Best for high-volume workloads, batch processing, and startups watching every token.

6

Models

$0.04

Avg / 1M tok

Mistral Small 3.1 Mistral $0.04 / 1M input

Phi-4 14B Microsoft $0.04 / 1M input

Gemma 3 12B Google $0.05 / 1M input

Qwen 3 30B Alibaba $0.05 / 1M input

Llama 3.1 8B Meta $0.02 / 1M input

Qwen 3 8B Alibaba $0.02 / 1M input