VerdictPal
Join

Models / Other

Hy3-preview (Reasoning)

Hy3-preview (Reasoning) · Released 2026-04-23

AA Intelligence Index 33.6. Listed on Artificial Analysis with API pricing $0.065/$0.235 per 1M tokens. Benchmark rows sync from the AA snapshot — see the ledger on this page.

Frontier model tracked on Artificial Analysis. Use the benchmark ledger below for sourced scores; we do not infer numbers beyond published rows.

#79 of 131 on AA Index · current snapshot
34 AA Index · 10 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite54
ContextUnlisted
Input / 1M$0.07
Output / 1M$0.23
Knowledge cutoffunpublished
Output speed166 t/s
TTFT (default API)2.01s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
87
Coding
41
Agentic
93
Long context
55

69average across 4 tested families

Benchmark placements

Where Hy3-preview (Reasoning) places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
AA output speedPerformance#17/ 149Artificial Analysis2026-07-28
τ³-Bench BankingAgentic#17/ 14393Artificial Analysis2026-07-28
AA time to first tokenPerformance#28/ 149Artificial Analysis2026-07-28
GPQA DiamondReasoning#64/ 14987Artificial Analysis2026-07-28
Humanity's Last ExamReasoning#73/ 14426Artificial Analysis2026-07-28
IFBenchReasoning#74/ 12163Artificial Analysis2026-07-28
SciCodeCoding#92/ 14441Artificial Analysis2026-07-28
Artificial Analysis Intelligence IndexReasoning#97/ 14734Artificial Analysis2026-07-28
AA-LCRLong context#114/ 14355Artificial Analysis2026-07-28
Terminal-BenchAgentic#115/ 14834Artificial Analysis2026-07-28
Family context

The three highest-scoring carded models in each capability family. Where Hy3-preview (Reasoning) shows up, it's highlighted.

ReasoningHy3-preview (Reasoning) · 87
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingHy3-preview (Reasoning) · 41
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticHy3-preview (Reasoning) · 93
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextHy3-preview (Reasoning) · 55
  1. 1
    GPT-5 (high)
    76
  2. 2
    GPT-5.2 Codex (xhigh)
    76
  3. 3
    Kimi K3
    75
Newest receipts

Benchmark rows added to the public ledger for Hy3-preview (Reasoning) in the last 120 days. Older rows live in the full table below.

10 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context54.67%Artificial Analysis6
AA output speedPerformance166 t/sArtificial Analysis6
AA time to first tokenPerformance2.01sArtificial Analysis6
Artificial Analysis Intelligence IndexReasoning33.6%Artificial Analysis6
GPQA DiamondReasoning86.7%Artificial Analysis6
Humanity's Last ExamReasoning25.5%Artificial Analysis6
IFBenchReasoning63.13%Artificial Analysis6
SciCodeCoding41.2%Artificial Analysis6
τ³-Bench BankingAgentic92.69%Artificial Analysis6
Terminal-BenchAgentic34.09%Artificial Analysis6
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning87
Coding41
Agentic93
Long context55
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotxAIMiniMaxMetaZhipuXiaomiOtherMeituanCohereNVIDIA

Full benchmark ledger

Every catalog benchmark for Hy3-preview (Reasoning). Scores link to the original source; gaps mean no public row exists yet.

10 sourced rows

10 of 32 catalog benchmarks have a sourced row for Hy3-preview (Reasoning).

Reasoning

Coding

Agentic

Long context

Performance

Sources