Pyyan / Compare / Hy-MT2-1.8B vs GPT-6.2 Astra

Hy-MT2-1.8B vs GPT-6.2 Astra

2 of 5

Language Models · verified 18 Nov 2026

Hy-MT2-1.8BTencentcurrent
GPT-6.2 AstraOpenAIcurrent
3 slots left
SpecificationHy-MT2-1.8BGPT-6.2 Astra
SummaryThe small sibling. Four cents per million tokens in.OpenAI's new frontier model, 20% cheaper per token than GPT-6 Astra.
Context8K1.05M
Input ($/Mtok)$0.044$8
Output ($/Mtok)$0.177$40
LicenceProprietaryProprietary
Max output8K128K
ParametersNot publishedNot published
Released20 Aug 202617 Nov 2026
Model IDhy-mt2-1.8bgpt-6.2-astra
CategoryLanguage ModelsLanguage Models
Official—OpenAI ↗

Highlighted rows are where these differ.

Hy-MT2-1.8B

  • The compact member of Tencent's MT2 translation pair
  • $0.044 in and $0.177 out per million tokens
  • Same 8K window as the larger model

Best for cheap bulk translation.

Full spec sheet →

GPT-6.2 Astra

  • Released on 17 November 2026 in ChatGPT Work, Codex and the API
  • $8 in, $0.40 cached and $40 out per million tokens, 20% below GPT-6 Astra on input and output
  • OpenAI says it leads Agents' Last Exam and Terminal-Bench-Science 0.1
  • OpenAI's first released model to meet its High threshold for AI self-improvement
  • Ultrafast mode available in the API at launch

Best for the hardest professional, coding and science work.

Full spec sheet →