DEEPSEEK API
DeepSeek V3.2 and R1, ready to call.
Call DeepSeek V3.2 for chat and code, or R1 for step-by-step reasoning, through one OpenAI-compatible endpoint and one API key.
Hand this to your agent
Add a Runcrate model to this project: run curl -fsSL"https://runcrate.ai/agents.md?intent=build"and follow it.
Your agent asks for an API key. Create a key
Quick start
Hand it to your agent. Or read the request it sends.
Your agent writes and runs it. The same key works for every model and every GPU.
Available models
Models you can call today. Billed prices, as charged.
| Model | Maker | Price | Detail |
|---|---|---|---|
| deepseek-ai/DeepSeek-V3.2 | DeepSeek | $0.273 in, $0.399 out per 1M tokens | 128K context, MoE architecture |
| deepseek-ai/DeepSeek-R1-0528 | DeepSeek | $0.525 in, $2.26 out per 1M tokens | Extended reasoning, chain-of-thought |
Why Runcrate
Built for production. One key, one balance.
Mixture of experts
DeepSeek V3.2 uses a Mixture-of-Experts architecture that activates only the parameters needed for each token, delivering frontier-quality output at lower compute cost.
Extended reasoning
DeepSeek R1 generates detailed chain-of-thought traces before answering, excelling at math olympiad problems, code debugging, and multi-step logic.
128K context window
Feed entire codebases, research papers, or lengthy documents into a single prompt without truncation.
Drop-in replacement
Uses the standard OpenAI chat completions format. Swap the base URL and model name, keep everything else.
Comparison
Runcrate and DeepSeek Platform, side by side.
| Feature | Runcrate | DeepSeek Platform |
|---|---|---|
| API format | OpenAI-compatible | OpenAI-compatible |
| Other models | Llama, Qwen, Mistral and more | DeepSeek only |
| Billing | One USD balance for every model | Separate DeepSeek account |
FAQ
Common questions.
Related