Early access: features, availability, and pricing may change as Kyara Intelligence evolves.
All models
Below baseline Reasoning

DeepSeek: DeepSeek V4.1 Flash

DeepSeek's first encoder-decoder Flash model with native vision and a 1M context.

Model ID deepseek/deepseek-v4.1-flash
Quick start

Drop this model into any OpenAI-compatible client by setting the base URL, your API key, and the model ID below.

curl https://api.kyara-intelligence.com/v1/chat/completions \
  -H "Authorization: Bearer $KYARA_INTELLIGENCE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4.1-flash",
    "messages": [
      { "role": "user", "content": "Hello!" }
    ]
  }'
See the full API reference
Overview

DeepSeek V4.1 Flash succeeds V4 Flash as DeepSeek's fast sparse MoE and is the first model built on its Causal Encoder-Decoder architecture, activating just 8B parameters on input and 16B on output from a 552B backbone to keep per-token compute low. Image understanding is native, with vision and text trained jointly from the start rather than added afterward. Configurable reasoning, a 1M-token context window, and up to 384K output tokens keep long conversations coherent while detailed replies arrive quickly at a low price per token.

Specifications
Context window
1M tokens Extreme
Cost vs catalog median
Input
0.8x
Output
1x
Input 5.4M credits / M tokens
Output 21.6M credits / M tokens

Multiples are relative to the catalog median; bars are scaled to the most expensive model.

Reasoning
Available
Status
Current
Moderation

Kyara Intelligence does not add a moderation layer. Moderating what goes in and out of your integration is up to you as the developer.

Kyara routes API calls without modifying inputs or outputs, and does not log prompt or response content. You are responsible for how your app uses the models and for complying with Kyara's terms.