Sapiens AIproprietary
Released: 2026-11-07Artificial AnalysisAgnes 2.5 Pro Theta
Agnes 2.5 Pro Theta from Sapiens AI, tracked on the Artificial Analysis Intelligence Index.
Intelligence Index
24.0/ 100
Artificial Analysis Intelligence Index (independent composite)Throughput Speed
197tok/sec
Output streaming throughputTime to First Answer
5.8s
Time to first answer token, including any thinkingCost to Run the Index
0.57USD
Artificial Analysis cost of the full Index runMeasured results · Artificial Analysis
Independent evaluations of Agnes 2.5 Pro Theta. Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.
- Intelligence Index
- 24.0
- Cost to run the Index
- $0.57
- Time per task
- 23.5 min
- Omniscience index
- -11.2
- Reasoning model
- Yes
- Cached-input price
- $0.01 / 1M
Reasoning & knowledge
- AA-Omniscience Accuracy21.3%
- CritPt15.1%
- CritPt v2 (Hard)26.0%
- HLE-Diamond16.8%
- Humanity's Last Exam37.2%
Coding & terminal agents
- FrontierCode-AA 2.016.6%
- SciCode56.0%
- Terminal-Bench 4.010.1%
- Terminal-Bench-Science 0.13.8%
Agentic work
- AA-Briefcase v1.1 (Elo)1496 Elo
- AutomationBench-AA48.8%
- GDPval-AA (normalized Elo)1507 Elo
Long context & documents
- AA-LCR (long-context reasoning)84.7%
- GDP.pdf14.2%
Multimodal & science
- MMMU-Pro78.4%
Health & safety
Headline Evaluation Breakdown
Six independently measured evaluations (Artificial Analysis)Reasoning · CritPt v2 (Hard)26.0%
Coding & agents · Terminal-Bench 4.010.1%
Long context · AA-LCR84.7%
Knowledge · Humanity's Last Exam37.2%
Sourced Benchmark Evaluations (14)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| Humanity's Last Exam | Active | 37.2% | 93 | 2026-11-15 | independent | Source ↗ |
| MMMU-Pro | Active | 78.4% | 93 | 2026-11-15 | independent | Source ↗ |
| SciCode | Active | 56% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 4.0 | Active | 10.1% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench-Science 0.1 | Active | 3.8% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-LCR (Long Context Reasoning) | Active | 84.7% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt | Active | 15.1% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt v2 (Hard) | Active | 26% | 93 | 2026-11-15 | independent | Source ↗ |
| HLE-Diamond | Active | 16.8% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Accuracy) | Active | 21.3% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Non-Hallucination) | Active | 58.6% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode-AA 2.0 | Active | 16.6% | 93 | 2026-11-15 | independent | Source ↗ |
| GDP.pdf | Active | 14.2% | 93 | 2026-11-15 | independent | Source ↗ |
| AutomationBench-AA | Active | 48.8% | 93 | 2026-11-15 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="sapiens-ai-agnes-2-5-pro-theta",
messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
temperature=0.2
)
print(response.choices[0].message.content)Evaluation & Sourcing Notes
Scores listed for Agnes 2.5 Pro Theta 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.