AI Benchmark Model
DeepSeek V4 Flash Vision (Reasoning, Max Effort)
Benchmark scores, pricing, speed, and model comparisons for DeepSeek V4 Flash Vision (Reasoning, Max Effort).
Key scores
DeepSeek V4 Flash Vision (Reasoning, Max Effort) is an artificial intelligence model developed by DeepSeek, officially released on August 21, 2026. The model demonstrates an intelligence index of 51.5 and a coding index of 65. In standardized testing, it achieved a GPQA score of 0.913, an HLE score of 0.345, a SciCode score of 0.466, and an LCR score of 0.78. Regarding operational performance, the model provides an output speed of 119.453 tokens per second with a time to first token of 0.746 seconds. Pricing for the model is set at $0.44 per million tokens for input and $1.32 per million tokens for output, resulting in a blended rate of $0.66 per million tokens.
Benchmark results
4 tracked benchmark rows for DeepSeek V4 Flash Vision (Reasoning, Max Effort).
GPQA Reasoning
91.3% LCR Other
78.0% SciCode Other
46.6% HLE Other
34.5%
Comments (0)
to join the discussion
No comments yet
Be the first to share your thoughts!