Pyyan / Compare / Hy4 flash vs Claude Fable 5.2

Hy4 flash vs Claude Fable 5.2

2 of 5

Language Models · verified 17 Nov 2026

Hy4 flashTencentcurrent
Claude Fable 5.2Anthropiccurrent
3 slots left
SpecificationHy4 flashClaude Fable 5.2
SummaryTencent's cheapest Hy model, 224B with 12B active, at $0.07 and $0.281 with a 1M context.Up to 30% faster than 5.1 and a fifth cheaper, with a fast mode at 2.5x.
Context1M1M
Input ($/Mtok)$0.07$8
Output ($/Mtok)$0.281$40
LicenceProprietary (weights promised)Proprietary
Max output64,000128K
Parameters224B total, 12B activeNot published
Released16 Nov 20269 Nov 2026
Model IDhy4-flashclaude-fable-5-2
CategoryLanguage ModelsLanguage Models
OfficialTencent ↗Anthropic ↗

Highlighted rows are where these differ.

Hy4 flash

  • Released on 16 November 2026 on Tencent Cloud TokenHub, WorkBuddy, Yuanbao, WeCom, QQ, ima and Marvis
  • 224B total parameters, including 64B Engram parameters, with 12B active per token
  • Reads text and images; supports reasoning, function calling, structured output and caching
  • Speaks the OpenAI Chat Completions, OpenAI Responses and Anthropic Messages APIs
  • List price $0.07 in and $0.281 out per million tokens, cache reads $0.003; half price on TokenHub until 30 November
  • Tencent says it plans to release the weights and that OpenRouter access is coming

Best for high-volume everyday office, document and agent work at a very low price.

Full spec sheet →

Claude Fable 5.2

  • Output up to 30% faster than Fable 5.1 at $8 and $40 per million tokens, against $10 and $50, with an optional fast mode up to 2.5x quicker at $16 and $80
  • 49.1% on FrontierCode 2.0 at top effort in Anthropic's table, against 42.8% for GPT-6 Astra
  • Same model as Mythos 5.2, which is limited to vetted cyber and life sciences organisations. Fable carries the cyber and biology safeguards
  • Anthropic says 1,000 agents formalised the proof of Fermat's Last Theorem in Lean in 30 hours, a result from its own prerelease testing

Best for the hardest long-running technical work.

Full spec sheet →