AI Benchmark Model
DeepSeek V4 Pro (Reasoning, High Effort)
Benchmark scores, pricing, speed, and model comparisons for DeepSeek V4 Pro (Reasoning, High Effort).
Key scores
DeepSeek V4 Pro is a reasoning-focused AI model released by DeepSeek on April 24, 2026. Designed for high-effort tasks, the model achieves an intelligence index of 43.7 and a coding index of 58.7. Performance metrics indicate an output speed of 61.151 tokens per second with a time to first token of 1.398 seconds. Pricing is structured at $0.43 per 1 million input tokens and $0.87 per 1 million output tokens, resulting in a blended rate of $0.54 per 1 million tokens. Benchmark results for the model include a GPQA score of 0.905, HLE of 0.352, SciCode of 0.464, IFBench of 0.713, LCR of 0.67, TerminalBench Hard of 0.417, and TAU2 of 0.942.
Benchmark results
7 tracked benchmark rows for DeepSeek V4 Pro (Reasoning, High Effort).
TAU2 Other
94.2% GPQA Reasoning
90.5% IFBench Other
71.3% LCR Other
67.0% SciCode Other
46.4% TerminalBench Hard Other
41.7% HLE Other
35.2%
Comments (0)
to join the discussion
No comments yet
Be the first to share your thoughts!