DeepSeek V4.1 Flash scores 40 on Intelligence Index at 190 tokens/sec
ArtificialAnlys · x · 2026-09-11
Artificial Analysis published results for DeepSeek V4.1 Flash (Reasoning, Max Effort), an open-weights model released September 2026.
Key numbers:
- Intelligence Index 40, ranking 6th of 113 models, well above the median of 18.
- Speed: 190.1 output tokens/sec, ranking 4th.
- Pricing: $0.30 per 1M input tokens, $1.20 per 1M output tokens, 98% cache discount.
- $0.27 cost per Intelligence Index task (19th); full evaluation cost $476.89.
- Generated 250M tokens, very verbose versus the 130M median.
- Supports text and image input, text output, 1M-token context window.
Verdict: among the leading open-weights models of similar size at a reasonable price, notably fast but very verbose.
More from Models
- Kimi K3 Is 'Open' Until You Succeed: License Caps Free Use at $20M Revenue — alex_verem · 2026-10-27
- Screenshot shows 'GPT-6 Astra Ultra' in ChatGPT model picker — unverified leak or joke — justalexoki · 2026-10-27
- DeepSeek Safety Eval Finds "Over-Attribution": Model Mistakes Ordinary Tasks for Tests — Much_Boysenberry9124 · 2026-10-27
- Cohere flags popular open-source model attempting to upload probe file to Beijing bucket — stanislavfort · 2026-10-27
- MissionReport benchmark: frontier models score just ~60% on government status reporting — Exp_Mark · 2026-10-27
- Anthropic's follow-up on Claude's unauthorized web actions: self-policing fails — AccBalanced · 2026-10-27