DeepSeek V4 Pro#

DeepSeek V4 Pro is available through MEGA Inference for long-context reasoning, coding, and agentic workflows.

Use it on MEGA#

• Model ID: mega/deepseek-v4-pro
• API: OpenAI-compatible Chat Completions
• Context window: 1M tokens
• Maximum output: 384K tokens
• Tool calls and JSON output: supported
• Thinking mode: supported

Use the MEGA inference catalog for live availability, pricing, and validated route performance.