VerdictPal
Join

Models / Other

Hy3-preview (Non-reasoning)

Hy3-preview (Non-reasoning) · Released 2026-04-23

AA Intelligence Index 26.1. 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.

#104 of 131 on AA Index · current snapshot
26 AA Index · 10 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite41
ContextUnlisted
Input / 1M$0.07
Output / 1M$0.23
Knowledge cutoffunpublished
Output speed182 t/s
TTFT (default API)2.07s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
73
Coding
39
Agentic
68
Long context
34

54average across 4 tested families

Benchmark placements

Where Hy3-preview (Non-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#16/ 149Artificial Analysis2026-07-28
AA time to first tokenPerformance#27/ 149Artificial Analysis2026-07-28
τ³-Bench BankingAgentic#47/ 14368Artificial Analysis2026-07-28
IFBenchReasoning#96/ 12148Artificial Analysis2026-07-28
SciCodeCoding#108/ 14439Artificial Analysis2026-07-28
Artificial Analysis Intelligence IndexReasoning#118/ 14726Artificial Analysis2026-07-28
Terminal-BenchAgentic#121/ 14832Artificial Analysis2026-07-28
AA-LCRLong context#128/ 14334Artificial Analysis2026-07-28
GPQA DiamondReasoning#130/ 14973Artificial Analysis2026-07-28
Humanity's Last ExamReasoning#132/ 1446Artificial Analysis2026-07-28
Family context

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

ReasoningHy3-preview (Non-reasoning) · 73
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingHy3-preview (Non-reasoning) · 39
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticHy3-preview (Non-reasoning) · 68
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextHy3-preview (Non-reasoning) · 34
  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 (Non-reasoning) in the last 120 days. Older rows live in the full table below.

10 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context33.67%Artificial Analysis4
AA output speedPerformance182 t/sArtificial Analysis4
AA time to first tokenPerformance2.07sArtificial Analysis4
Artificial Analysis Intelligence IndexReasoning26.1%Artificial Analysis4
GPQA DiamondReasoning73.2%Artificial Analysis4
Humanity's Last ExamReasoning6.3%Artificial Analysis4
IFBenchReasoning47.96%Artificial Analysis4
SciCodeCoding39.4%Artificial Analysis4
τ³-Bench BankingAgentic67.54%Artificial Analysis4
Terminal-BenchAgentic31.82%Artificial Analysis4
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.

Reasoning73
Coding39
Agentic68
Long context34
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 (Non-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 (Non-reasoning).

Reasoning

Coding

Agentic

Long context

Performance

Sources