Gateway operational · 16 models live · GPU capacity on demand Read the docs →
LINGYUNS gateway
Open weights · DeepSeek

DeepSeek-V3

671B MoE open model with strong general and coding performance.

Context window 64K tokens
Input price $0.27 per 1M tokens
Output price $1.1 per 1M tokens
Routing Included failover + retries

Call it with any OpenAI SDK

Swap the base URL and key — request shape, streaming and tool calls stay identical.

python
from openai import OpenAI

client = OpenAI(
    base_url="https://api.lingyuns.com/v1",
    api_key="ly-sk-…",
)

r = client.chat.completions.create(
    model="deepseek-v3",
    messages=[{"role": "user", "content": "Hello"}],
)
# billed at $0.27 in / $1.1 out per 1M
Full quickstart

Specifications

Model IDdeepseek-v3
ProviderDeepSeek
CategoryOpen weights
Context window64K tokens
Input / 1M$0.27
Output / 1M$1.1
ModalityText · streaming · tools

Prices are list rates in USD per million tokens. Your prepaid credits cover every model in the catalog at these rates.

Alternatives

Other Open weights models

Llama 3.1 405B

The largest open-weights model — GPT-4-class quality you can self-host later.

Context128K
Input / 1M$2.7
Output / 1M$2.7

Llama 3.1 70B

The open-weights workhorse for fine-tuning and private deployments.

Context128K
Input / 1M$0.52
Output / 1M$0.75

Llama 3.1 8B

Ultra-cheap 8B model for classification, routing and simple extraction.

Context128K
Input / 1M$0.05
Output / 1M$0.08