deepseek-v4-flash-vision-exp-dspark-1m DeepSeek V4 Flash with image input and a 1M-token context — reasoning, tool calling, and structured outputs behind an OpenAI-compatible endpoint.
No GPU. No setup. Cancel anytime.
DeepSeek V4 Flash + Vision is the long-context model of Sogni Intelligence: DeepSeek's V4 Flash with a 1M-token context window, reasoning, native tool calling, structured outputs, and image input, served through the same OpenAI-compatible endpoint as Qwen3.6. Point an OpenAI SDK at api.sogni.ai and pass model id deepseek-v4-flash-vision-exp-dspark-1m.
It already powers Sogni's own tools. Sogni Chat's Unrestricted and Thinking + Unrestricted modes run on it, and Sogni Web uses it as the Prompt Assistant and AI Screenwriter for Unlimited members, where its broader world knowledge turns short ideas into richer, more faithful prompts and video scripts.
Choose it when a job needs a very large context, such as whole codebases, long transcripts, or document sets, or when a request mixes text and images. Qwen3.6 35B-A3B stays the default for everyday chat at a lower output price. Vision support is experimental, as the model id's Vision Exp label indicates.
Use pay-as-you-go Spark packs for each request (1 Spark = $0.005), or choose a flat-rate Sogni plan for credit-free fair-use generation in the app.
| Configuration | Spark | USD |
|---|---|---|
| 1M input tokens | 70.0 Spark | $0.35 |
| 1M output tokens | 550 Spark | $2.75 |
| 10K in + 2K out (typical request) | 1.80 Spark | $0.009 |
DeepSeek V4 Flash + Vision needs an Unlimited plan, Premium Spark, or SOGNI. Free Spark cannot pay for it.
One Sogni API key reaches every model on the Supernet — call DeepSeek V4 Flash + Vision with the exact model id.
const res = await fetch('https://api.sogni.ai/v1/chat/completions', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
Authorization: `Bearer ${process.env.SOGNI_API_KEY}`,
},
body: JSON.stringify({
model: 'deepseek-v4-flash-vision-exp-dspark-1m',
messages: [{ role: 'user', content: 'Write a haiku about decentralized GPUs.' }],
}),
});
const { choices } = await res.json();
console.log(choices[0].message.content);curl https://api.sogni.ai/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $SOGNI_API_KEY" \
-d '{
"model": "deepseek-v4-flash-vision-exp-dspark-1m",
"messages": [{ "role": "user", "content": "Write a haiku about decentralized GPUs." }]
}' OpenAI-compatible — point any OpenAI client at https://api.sogni.ai/v1. Full reference at docs.sogni.ai.
Use a flat monthly plan for credit-free fair-use generation, or buy Spark packs when pay-as-you-go fits better. Both run on the same creator-owned GPU network.
One flat price in the app. Generate under fair use without a per-image meter.
Image, video, music, and language models in one workspace and one API key.
Prefer pay-as-you-go? Call DeepSeek V4 Flash + Vision by id and pay with Spark packs.
Runs on a decentralized GPU network where workers share subscription revenue.
It is DeepSeek's V4 Flash language model served on the Sogni Supernet with image input, a 1M-token context window, reasoning, tool calling, and structured outputs. You call it through Sogni Intelligence's OpenAI-compatible API with model id deepseek-v4-flash-vision-exp-dspark-1m.
Pay-as-you-go pricing is 70 Spark ($0.35) per 1M input tokens and 550 Spark ($2.75) per 1M output tokens, so a typical request with 10K input and 2K output tokens costs 1.8 Spark ($0.009). It is also included under fair use on Sogni Unlimited and Unlimited Pro.
No. DeepSeek V4 Flash + Vision needs an Unlimited plan, Premium Spark, or SOGNI. Qwen3.6 35B-A3B remains available for everyday chat.
The context window is 1,048,576 tokens. Each request can return up to 32,768 output tokens, with 8,192 by default.
Yes. Send images as image_url content parts in an OpenAI-compatible chat completions request, alongside your text. Vision support is experimental.
Sogni Chat runs its Unrestricted and Thinking + Unrestricted modes on it, and Sogni Web uses it as the Prompt Assistant and AI Screenwriter for Unlimited members. Developers can call it directly through the Sogni API.
Use Qwen3.6 35B-A3B, the default model, for everyday chat, coding, and tool calling with a 262K-token context at $0.90 per 1M output tokens. Use DeepSeek V4 Flash + Vision when you need a 1M-token context or broader world knowledge.
No. DeepSeek V4 Flash + Vision runs on the Sogni Supernet, a network of creator GPUs, with no local install or graphics card required.
Create in the app, or build with the API. Your call.