Tencentopen-weight
Released: 2024-11-05

Hunyuan-Large

Tencent largest open-weight Mixture-of-Experts architecture with 389B parameters and 256K context.

Intelligence Index
84.5/ 100
Calibrated multi-domain composite
Throughput Speed
65tok/sec
Output streaming throughput
Time To First Token
410ms
Initial chunk server latency
LiveBench Contam-Free
74.8%
September 2026 suite score

Sourced Benchmark Evaluations (2)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
ARC-AGI-2Nearing Saturation64.7%
75
2024-12-03independentSource ↗
SWE-bench VerifiedSaturated55.5%
25
2024-11-14independentSource ↗
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run hunyuan-large

# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "hunyuan-large",
    "messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
  }'

Evaluation & Sourcing Notes

Scores listed for Hunyuan-Large represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.