SI Index Data as of Oct 8, 2026

Hy3: Price, Context, Benchmarks, and Release Details

tencent provisional listing open weights · first seen Oct 8, 2026 · score computed Oct 8, 2026, 22:13 UTC
60.7
SI Score (method si-v2-absolute-shrinkage-1)
#57 of 115 ranked
88% confidence 88 percent, High confidence — 2 of 3 expected sources in
Coverage 70% of expected source weight · 100% confidence at 80% coverage

Pillar breakdown

Coding (weight 40 percent) 69.2
Math (weight 15 percent) —
Preference (weight 15 percent) 76.9
Reasoning (weight 30 percent) 54.2

Pillar weights: reasoning 30% · math 15% · coding 40% · preference 15%. Benchmark results use fixed 0–100 scales before averaging; incomplete evidence is shrunk toward 50.

Facts

Input price / 1M
not yet reported
Output price / 1M
not yet reported
Context window
256Kmodels.devPublished source fact Retrieved Oct 8, 2026 · MIT
Open source ↗
Max output
128Kmodels.devPublished source fact Retrieved Oct 8, 2026 · MIT
Open source ↗
Released
Jul 6, 2026models.devPublished source fact Retrieved Oct 8, 2026 · MIT
Open source ↗
Open weights
Yesmodels.devPublished source fact Retrieved Oct 8, 2026 · MIT
Open source ↗
License
not yet reported
Input modalities
textmodels.devPublished source fact Retrieved Oct 8, 2026 · MIT
Open source ↗

Benchmark results

7 results
Benchmark Raw result Normalized (0–100) Pillar
gpqa diamond 90.4%Official model cards via models.devLab-reported; metric score; transcribed by MIT models.dev catalog; not independently evaluated [variant] highest reasoning effort Retrieved Oct 8, 2026 · factual citation; MIT transcription
Open source ↗ 		90.4 reasoning
hle text only 37%Official model cards via models.devLab-reported; metric score; transcribed by MIT models.dev catalog; not independently evaluated [variant] highest reasoning effort; without tools; text-only Retrieved Oct 8, 2026 · factual citation; MIT transcription
Open source ↗ 		37.0 reasoning
hle text only tools 53.2%Official model cards via models.devLab-reported; metric score; transcribed by MIT models.dev catalog; not independently evaluated [variant] highest reasoning effort; with tools; text-only Retrieved Oct 8, 2026 · factual citation; MIT transcription
Open source ↗ 		53.2 reasoning
lmarena text 1440.7 eloLMArena / ArenaPublished source fact [variant] text / overallPublished Oct 2, 2026 Retrieved Oct 8, 2026 · CC-BY-4.0
Open source ↗ 		76.9 preference
swe bench pro 57.9%Official model cards via models.devLab-reported; metric score; transcribed by MIT models.dev catalog; not independently evaluated [variant] highest reasoning effort; SWE-agent Retrieved Oct 8, 2026 · factual citation; MIT transcription
Open source ↗ 		57.9 coding
swe bench verified 78%Official model cards via models.devLab-reported; metric resolved; transcribed by MIT models.dev catalog; not independently evaluated [variant] Retrieved Oct 8, 2026 · factual citation; MIT transcription
Open source ↗ 		78.0 coding
terminal bench v2 1 71.7%Official model cards via models.devLab-reported; metric score; transcribed by MIT models.dev catalog; not independently evaluated [variant] highest reasoning effort; 4h timeout; 500 episodes; Terminus 2; 2.1 Retrieved Oct 8, 2026 · factual citation; MIT transcription
Open source ↗ 		71.7 coding

Normalization uses fixed absolute 0–100 scales for each unit, independently of other models that have a result on each benchmark. Coverage and evidence breadth still affect the composite; compare the evaluation conditions before reading a small score gap as decisive. “lab-reported” marks the provider's own published figure.

Sources: in and pending

in Reported (2)

  • Official model cards via models.dev · arrived Oct 8, 2026
  • LMArena / Arena · arrived Oct 8, 2026

pending Awaiting (1)

  • LiveBench · carries 30% of expected weight

The confidence % rises as pending sources publish. Some sources never cover some models — that is why 100% confidence arrives at 80% of expected weight, not at full coverage.