readable AI benchmarks (simplified)
Trillion Labs
Tri-21B-Think
Tri-21B-Think has a Quality Score of 16.3, ranking 475th among 571 scored models. It has a 32k token context window. Its Reliability Score is 18.7, ranking 495th among 573 scored models, below average (total average: 36.5).
Quality Score16.3coverage 100.0%: AA Intelligence and AA-Omniscience outcomes available
Value Scorenot available
Factual reliability18.7
Cache Discountnot available
Model specification
- Reasoning
- reasoning
- Input modalities
- none
- Output modalities
- none
- Context window
- 32k tokens
- Weights
- open weights
Token prices USD per 1M tokens
- Input
- not available
- Output
- not available
- Cached Input
- not available
- Cache write
- not available
- Cost per task
- not available
Capability
- Intelligence
- 9.0
- Coding
- not available
- Agentic
- not available
- Omniscience
- -62.5
- Correct
- 12.3%
- Blended price
- not available
Answer outcomes
Fully graded outcomes (Correct + Incorrect): 87.2%
Correct 12.3%Incorrect 74.9%Partial / not attempted 12.8%
Artificial Analysis benchmarks
- Humanity’s Last Exam
- 6%
- GPQA Diamond
- 60%
- CritPt
- 0%
- AA-Omniscience
- 19%
- AA-LCR
- 16%
Similar models
- Tri-21B-think PreviewTrillion LabsQuality Score 16.8Value Score not available
- gpt-oss-20b (High)OpenAIQuality Score 16.1Value Score not available
- EXAONE 4.0 32B (Reasoning)LG AI ResearchQuality Score 16.3Value Score not available
- Motif-2-12.7B-ReasoningMotif TechnologiesQuality Score 16.6Value Score not available
- Qwen3 4B 2507 (Reasoning)AlibabaQuality Score 16.7Value Score not available
- Sarvam 105B (High)SarvamQuality Score 16.9Value Score 16.0