deepseek's new ai model costs 3 cents to run. claude's flagship costs $3.15 for the same job

V4-Flash is more than 100x cheaper than Anthropic's Claude Fable 5 on the same benchmark — here's what that price gap actually buys you, and what it… Aliteq Lena Fischer · AI & Local Compute Editor

DeepSeek's V4-Flash charges $0.14 per million input tokens and $0.28 per million output tokens — a fraction of frontier-model pricing.

The numbers that matter

Artificial Analysis measured an average cost of 3 cents per benchmark task for V4-Flash, versus $3.15 for Claude Fable 5.

The numbers that matter

V4-Flash scores 50/100 on the Intelligence Index, tying Gemini 3.6 Flash but trailing Claude Opus 5 and GPT-5.6 by roughly nine points.

The numbers that matter

It's a mixture-of-experts model — 304B parameters on Hugging Face, including a speculative-decoding draft module on top of a ~284B base — released under an MIT license.

The numbers that matter

If your workload is high-volume and tolerant of a mid-tier reasoning ceiling — support triage, tagging, first-draft copy — V4-Flash's price makes it close to a no-brainer to test. If you're running…

Read the full story

Aliteq deepseek's new ai model costs 3 cents to run. claude's flagship costs $3.15 for the same job Read the full story on Aliteq