Open-weight models, serverless
All models use OpenAI-compatible IDs. Drop them into your existing SDK and start paying per token — no GPU provisioning needed.
// Works with any OpenAI-compatible SDK
const client = new OpenAI({
baseURL: "https://api.lexora.network/v1",
apiKey: process.env.LEXORA_API_KEY,
});
// Chat completion
await client.chat.completions.create({
model: "Qwen/Qwen3-8B", // ← use any model ID below
messages: [{ role: "user", content: "Hello!" }],
});
// Image generation
await client.images.generate({
model: "black-forest-labs/FLUX.1-schnell",
prompt: "A futuristic city at night",
});Run directly on Lexora's distributed worker network. Flat per-token pricing — no markup, no plan differences.
Qwen3 8B
Qwen/Qwen3-8B
Fast reasoning model on the network. Optimized for chat, code, classification, and high-throughput integrations.
Common use cases
FLUX.1 schnell
black-forest-labs/FLUX.1-schnell
Guidance-free 4-step diffusion model. Best speed-to-quality ratio for programmatic image generation at scale.
Common use cases
Llama 3.1 8B Instruct
meta-llama/Llama-3.1-8B-Instruct
Balanced performance and cost. Strong on reasoning, code, and complex instruction following.
Llama 3.1 70B Instruct
meta-llama/Llama-3.1-70B-Instruct
Flagship open-weight LLM. Near-GPT-4 quality for complex reasoning and long-context tasks.
Stable Diffusion XL
stabilityai/stable-diffusion-xl-base-1.0
Industry-standard image model with a huge ecosystem of LoRAs, ControlNets, and refiners.
Served via provider partnerships. Pricing is market rate plus a small markup to cover routing and support.
DeepSeek V4 Flash
deepseek/deepseek-v4-flash
per 1M tokens · Pro
Fast variant of DeepSeek V4. High-throughput reasoning at low cost, ideal for production pipelines.
DeepSeek V4 Pro
deepseek/deepseek-v4-pro
per 1M tokens · Pro
Full DeepSeek V4. Top-tier reasoning and coding performance, competitive with frontier models.
Kimi K2.6
moonshot/kimi-k2.6
per 1M tokens · Pro
Moonshot's flagship long-context model. Exceptional multilingual reasoning and document understanding.
Kimi K2.7 Code
moonshot/kimi-k2.7-code
per 1M tokens · Pro
Code-specialized Kimi model. Strong at code generation, debugging, and technical documentation.
Kimi K2.7 Code Highspeed
moonshot/kimi-k2.7-code-highspeed
per 1M tokens · Pro
High-throughput variant of Kimi K2.7 Code. Lower latency for latency-sensitive coding pipelines.
FLUX.1 dev
black-forest-labs/FLUX.1-dev
Guidance-distilled FLUX for higher prompt adherence and richer detail. Tunable CFG scale.
Partnership model prices shown at Pro rate (+2%). Pay As You Go users are charged +10% on top of market rate. Upgrade to Pro →
Need a specific model?
We add models based on demand. If you need a model that isn't listed — a specific Mistral variant, a multimodal model, a code model, or a custom deployment — let us know and we'll prioritize it.
Request a model →