PyyanThe AI Index

Pyyan / Compare / Inkling vs DeepSeek V4-Pro

Inkling vs DeepSeek V4-Pro

2 of 5

Open-Weight Models · verified 13 Sept 2026

TInklingThinking Machines Labcurrent
DeepSeek V4-ProDeepSeekcurrent
3 slots left
SpecificationInklingDeepSeek V4-Pro
Summary975B under Apache 2.0, shipped as a starting point rather than a finished product.Best all-round open model of 2026.
Context1M—
Input ($/Mtok)Weights only—
Output ($/Mtok)Weights only—
LicenceApache 2.0MIT
Max outputNot published—
Parameters975B total, 41B active—
Released15 Jul 2026—
Model IDNot published—
CategoryOpen-Weight ModelsOpen-Weight Models
OfficialThinking Machines Lab ↗DeepSeek ↗

Highlighted rows are where these differ.

Inkling

  • Pretrained on 45 trillion tokens of text, images, audio and video, with an encoder free architecture
  • Apache 2.0, weights on Hugging Face, same day fine tuning through the lab's Tinker platform
  • The lab earns from the compute customers use to adapt it, not from per token access

Best for fine tuning on your own data.

Full spec sheet →

DeepSeek V4-Pro

  • Tops open leaderboards on agentic coding and reasoning
  • MIT licence — no conditions
  • Pioneered aggressive cache-hit pricing at ~$0.07/M
  • DeepSeek routes all V4-Pro requests to V4.1 Flash from 14 September 2026

Best for everything open.

Full spec sheet →