CoRover / BharatGPTproprietary
Released: 2024-10-18BharatGPT-14B
Sovereign multimodal enterprise AI model engineered for citizen engagement, banking, railway ticketing, and governance across 14+ Indian languages.
Intelligence Index
81.2/ 100
Calibrated multi-domain compositeThroughput Speed
110tok/sec
Output streaming throughputTime To First Token
230ms
Initial chunk server latencyLiveBench Contam-Free
70.8%
September 2026 suite scoreSourced Benchmark Evaluations (61)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| LiveBench (Continuous) | Active | 70.8% | 94 | 2025-02-01 | independent | Source ↗ |
| ARC-AGI-3 | Active | 51.7% | 93 | 2024-11-22 | independent | Source ↗ |
| ARC-AGI-2 | Nearing Saturation | 62% | 75 | 2024-11-15 | independent | Source ↗ |
| FrontierMath | Active | 21% | 98 | 2024-11-12 | independent | Source ↗ |
| PutnamBench | Active | 39% | 90 | 2024-11-11 | independent | Source ↗ |
| Omni-MATH | Deprecated | 55.5% | 10 | 2024-11-09 | independent | Source ↗ |
| SWE-Lancer | Deprecated | 47.6% | 20 | 2024-11-09 | independent | Source ↗ |
| Soohak | Active | 57% | 98 | 2024-11-07 | independent | Source ↗ |
| FrontierCode | Active | 40.6% | 98 | 2024-11-07 | independent | Source ↗ |
| Humanity's Last Exam | Active | 35% | 93 | 2024-11-05 | independent | Source ↗ |
| StructFlowBench | Nearing Saturation | 63.5% | 75 | 2024-11-05 | independent | Source ↗ |
| AIME | Nearing Saturation | 71.3% | 75 | 2024-11-05 | independent | Source ↗ |
| OlympiadBench | Active | 61.5% | 80 | 2024-11-04 | independent | Source ↗ |
| FireBench | Active | 62.8% | 95 | 2024-11-03 | independent | Source ↗ |
| Terminal-Bench | Active | 49% | 80 | 2024-11-03 | independent | Source ↗ |
| MathArena | Active | 60% | 99 | 2024-11-03 | independent | Source ↗ |
| OpenAI-MRCR | Active | 68.2% | 80 | 2024-11-02 | independent | Source ↗ |
| CodeContests | Deprecated | 45.5% | 10 | 2024-11-02 | independent | Source ↗ |
| SimpleQA | Active | 42.2% | 90 | 2024-11-02 | independent | Source ↗ |
| ComplexBench | Nearing Saturation | 61.3% | 80 | 2024-11-01 | independent | Source ↗ |
| SWE-bench Pro | Active | 43.4% | 90 | 2024-11-01 | independent | Source ↗ |
| LongBench-Pro | Active | 66.6% | 80 | 2024-11-01 | independent | Source ↗ |
| MultiPL-E | Deprecated | 64.4% | 10 | 2024-10-31 | independent | Source ↗ |
| ZeroSCROLLS | Nearing Saturation | 65% | 78 | 2024-10-31 | independent | Source ↗ |
| AgentIF | Nearing Saturation | 50.8% | 80 | 2024-10-30 | independent | Source ↗ |
| MATH | Saturated | 73.5% | 25 | 2024-10-30 | independent | Source ↗ |
| LiveCodeBench | Saturated | 57.4% | 39 | 2024-10-30 | independent | Source ↗ |
| InfiniteBench | Nearing Saturation | 69% | 65 | 2024-10-30 | independent | Source ↗ |
| MMLU-Pro | Nearing Saturation | 74.7% | 75 | 2024-10-29 | independent | Source ↗ |
| Mercury | Deprecated | 61.6% | 10 | 2024-10-29 | independent | Source ↗ |
| SCROLLS | Nearing Saturation | 65% | 65 | 2024-10-29 | independent | Source ↗ |
| RULER | Nearing Saturation | 74.7% | 84 | 2024-10-28 | independent | Source ↗ |
| TabMWP | Saturated | 69% | 25 | 2024-10-28 | independent | Source ↗ |
| CyberSecEval | Deprecated | 63% | 10 | 2024-10-28 | independent | Source ↗ |
| MMLU-Redux | Saturated | 79.6% | 25 | 2024-10-27 | independent | Source ↗ |
| QuALITY | Nearing Saturation | 71.5% | 78 | 2024-10-27 | independent | Source ↗ |
| SWE-bench Verified | Saturated | 54.6% | 25 | 2024-10-27 | independent | Source ↗ |
| MGSM | Nearing Saturation | 71.3% | 65 | 2024-10-26 | independent | Source ↗ |
| LongBench | Nearing Saturation | 69.8% | 65 | 2024-10-26 | independent | Source ↗ |
| SWE-bench | Deprecated | 47.6% | 10 | 2024-10-26 | independent | Source ↗ |
| AGIEval | Saturated | 68.4% | 25 | 2024-10-25 | independent | Source ↗ |
| QASPER | Nearing Saturation | 67.4% | 65 | 2024-10-25 | independent | Source ↗ |
| TruthfulQA | Saturated | 71.5% | 25 | 2024-10-25 | independent | Source ↗ |
| SuperGLUE | Saturated | 81.1% | 25 | 2024-10-24 | independent | Source ↗ |
| GSM-Hard | Deprecated | 72% | 10 | 2024-10-24 | independent | Source ↗ |
| MT-Bench | Nearing Saturation | 7.7% | 75 | 2024-10-23 | independent | Source ↗ |
| BIG-Bench Hard | Saturated | 77.9% | 25 | 2024-10-23 | independent | Source ↗ |
| CodeXGLUE | Deprecated | 65.8% | 23 | 2024-10-23 | independent | Source ↗ |
| MBPP+ | Nearing Saturation | 71.4% | 65 | 2024-10-22 | independent | Source ↗ |
| WinoGrande | Saturated | 82.7% | 28 | 2024-10-22 | independent | Source ↗ |
| IFEval | Deprecated | 73.2% | 25 | 2024-10-22 | independent | Source ↗ |
| GPQA | Nearing Saturation | 71.5% | 78 | 2024-10-22 | independent | Source ↗ |
| HumanEval+ | Saturated | 75.6% | 25 | 2024-10-21 | independent | Source ↗ |
| Needle-in-a-Haystack | Saturated | 96.2% | 44 | 2024-10-21 | independent | Source ↗ |
| CommonsenseQA | Saturated | 83.5% | 28 | 2024-10-21 | independent | Source ↗ |
| GSM8K | Saturated | 81% | 25 | 2024-10-20 | independent | Source ↗ |
| GLUE | Saturated | 85.9% | 28 | 2024-10-20 | independent | Source ↗ |
| MBPP | Saturated | 74.2% | 25 | 2024-10-20 | independent | Source ↗ |
| HumanEval | Saturated | 79.8% | 25 | 2024-10-19 | independent | Source ↗ |
| MMLU | Saturated | 82.8% | 25 | 2024-10-19 | independent | Source ↗ |
| ARC-AGI | Saturated | 74.7% | 25 | 2024-10-19 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="bharatgpt-14b",
messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
temperature=0.2
)
print(response.choices[0].message.content)Evaluation & Sourcing Notes
Scores listed for BharatGPT-14B 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.