Catalog
DeepSeek
DeepSeek V4 Flash
deepseek/deepseek-v4-flash
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window.
- Input
- $0.14 / 1M
- Output
- $0.28 / 1M
- Context
- 1M
- Max output
- 131.1K
- Released
- Apr 24, 2026
- Providers
- 1 live of 4
Providers
Requests go to your organization's endpoints first, then the default host, then fall back in order. Pin hosts with provider.order or provider.only.
| Provider | Region | Input / 1M | Output / 1M | Cache read | Context | Max output | Route |
|---|---|---|---|---|---|---|---|
Fireworks Plural | us | $0.14 / 1M | $0.28 / 1M | — | 1M | 131.1K | Default |
Azure Plural | us | $0.21 / 1M | $0.56 / 1M | $0.031 / 1M | 1M | 131.1K | Unconfigured |
Baseten Plural | us | $0.16 / 1M | $0.32 / 1M | — | 1M | 131.1K | Unconfigured |
DeepSeek Plural | cn | $0.14 / 1M | $0.28 / 1M | — | 1M | 131.1K | Unconfigured |
Use this model
One Plural API key reaches every provider above, with fallback and tracing built in. Create a key if you don't have one.
Sign in with your API key
Terminal plural auth login --api-key-stdinCall DeepSeek V4 Flash
CLI # Run it on a Task in your projectplural run --task my-task --model deepseek/deepseek-v4-flash # Or save it as an Agent with its own instructionsplural agent init my-agent --model deepseek/deepseek-v4-flashplural run --task my-task --agent my-agent
Using another client? The gateway is OpenAI-compatible at https://api.pluralintel.com/v1.