Hy3 Preview

Hy3 Preview is a reasoning model from Tencent in the Hy3 family. 16 benchmarks count toward its score, in 5 categories.

availableShows if the model has enough results for an index.
IndexOverall score. 50 is the middle.55.6 ±6.5
CoverageShare of the index weight with results.75%
SpeedOutput tokens per second.86/s
Input / 1MUS dollars per 1M input tokens.$0.18
Output / 1MUS dollars per 1M output tokens.$0.6
ContextMaximum tokens in one request.262K
EloLMArena rating and rank.N/A

50 is the middle of the board. The range shows the doubt in the index.

CapabilitiesScore per category. 50 is the middle.

50 is the middle
AgenticMulti-step tasks with tools.
55.5
CodingCode writing and repair.
57.3
ReasoningLogic problems and puzzles.
51.5
MultimodalTasks with images and text.
N/A
KnowledgeFacts and expert knowledge.
59.5
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
39.5
MathMath problems.
N/A

Results

16 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result on the index scale.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
Artificial Analysis GPQA DiamondKnowledge89.7%60.5Artificial Analysis
Graduate-Level Google-Proof Q&AKnowledge87.2%58.7David Rein et al.
GPQA DiamondKnowledge87.2%58.7David Rein et al.
Software Engineering Benchmark VerifiedCoding74.4%57.2Carlos E. Jimenez et al.
Artificial Analysis Long Context ReasoningReasoning66.7%54.4Artificial Analysis
Instruction Following BenchmarkInstruction63.1%39.5Benchmark authors
Artificial Analysis Coding IndexCoding58.8%60.4Artificial Analysis
Artificial Analysis SciCodeCoding48.6%60.3Artificial Analysis
Artificial Analysis Intelligence IndexKnowledge41.2%74.0Artificial Analysis
Scientific Code BenchmarkCoding41.2%51.2Benchmark authors
Gert Labs Composite Game BenchmarkAgentic36.9%45.3Gert Labs
GDPval-AA normalizedAgentic35.8%63.1Artificial Analysis
Artificial Analysis Humanity's Last ExamKnowledge33.5%60.9Artificial Analysis
Artificial Analysis Omniscience AccuracyKnowledge31.5%52.8Artificial Analysis
Artificial Analysis Agentic IndexAgentic25.6%58.1Artificial Analysis
Humanity's Last ExamKnowledge25.5%50.4Center for AI Safety et al.
Critical Physics TasksReasoning4.9%48.7Artificial Analysis

16 benchmarks count, from 17 of 17 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authorsOpenRouter, collected directlyNo licence stated

Same level, lower price

Qwen3.8-Flash-Next66.4 · Freedots3-note Preview65.5 · FreeApodex 1.1 Mini59.1 · Free

More from Tencent

Hy4 preview70.2Hy357.5AuKUnrankedHunyuan A13B InstructUnrankedHy ASR 3.0 previewUnranked