Hanooman AI (SML)open-weight
Released: 2024-05-15Hanooman 7B
Open multimodal Indic model built in partnership with Indian engineering institutes for localized sectoral solutions.
Intelligence Index
79.0/ 100
Calibrated multi-domain compositeThroughput Speed
125tok/sec
Output streaming throughputTime To First Token
210ms
Initial chunk server latencyLiveBench Contam-Free
68.5%
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 | 68.5% | 94 | 2025-02-01 | independent | Source ↗ |
| ARC-AGI-3 | Active | 50.1% | 93 | 2024-06-19 | independent | Source ↗ |
| ARC-AGI-2 | Nearing Saturation | 60.1% | 75 | 2024-06-12 | independent | Source ↗ |
| FrontierMath | Active | 20.4% | 98 | 2024-06-09 | independent | Source ↗ |
| PutnamBench | Active | 38% | 90 | 2024-06-08 | independent | Source ↗ |
| Omni-MATH | Deprecated | 54% | 10 | 2024-06-06 | independent | Source ↗ |
| SWE-Lancer | Deprecated | 46.2% | 20 | 2024-06-06 | independent | Source ↗ |
| Soohak | Active | 55.5% | 98 | 2024-06-04 | independent | Source ↗ |
| FrontierCode | Active | 39.4% | 98 | 2024-06-04 | independent | Source ↗ |
| Humanity's Last Exam | Active | 33.9% | 93 | 2024-06-02 | independent | Source ↗ |
| StructFlowBench | Nearing Saturation | 61.8% | 75 | 2024-06-02 | independent | Source ↗ |
| AIME | Nearing Saturation | 69.4% | 75 | 2024-06-02 | independent | Source ↗ |
| OlympiadBench | Active | 59.9% | 80 | 2024-06-01 | independent | Source ↗ |
| FireBench | Active | 61.1% | 95 | 2024-05-31 | independent | Source ↗ |
| Terminal-Bench | Active | 47.6% | 80 | 2024-05-31 | independent | Source ↗ |
| MathArena | Active | 58.4% | 99 | 2024-05-31 | independent | Source ↗ |
| OpenAI-MRCR | Active | 66.4% | 80 | 2024-05-30 | independent | Source ↗ |
| CodeContests | Deprecated | 44.2% | 10 | 2024-05-30 | independent | Source ↗ |
| SimpleQA | Active | 41.1% | 90 | 2024-05-30 | independent | Source ↗ |
| ComplexBench | Nearing Saturation | 59.6% | 80 | 2024-05-29 | independent | Source ↗ |
| SWE-bench Pro | Active | 42.2% | 90 | 2024-05-29 | independent | Source ↗ |
| LongBench-Pro | Active | 64.8% | 80 | 2024-05-29 | independent | Source ↗ |
| MultiPL-E | Deprecated | 62.6% | 10 | 2024-05-28 | independent | Source ↗ |
| ZeroSCROLLS | Nearing Saturation | 63.2% | 78 | 2024-05-28 | independent | Source ↗ |
| AgentIF | Nearing Saturation | 49.4% | 80 | 2024-05-27 | independent | Source ↗ |
| MATH | Saturated | 71.5% | 25 | 2024-05-27 | independent | Source ↗ |
| LiveCodeBench | Saturated | 55.8% | 39 | 2024-05-27 | independent | Source ↗ |
| InfiniteBench | Nearing Saturation | 67.1% | 65 | 2024-05-27 | independent | Source ↗ |
| MMLU-Pro | Nearing Saturation | 72.7% | 75 | 2024-05-26 | independent | Source ↗ |
| Mercury | Deprecated | 59.8% | 10 | 2024-05-26 | independent | Source ↗ |
| SCROLLS | Nearing Saturation | 63.2% | 65 | 2024-05-26 | independent | Source ↗ |
| RULER | Nearing Saturation | 72.7% | 84 | 2024-05-25 | independent | Source ↗ |
| TabMWP | Saturated | 67.2% | 25 | 2024-05-25 | independent | Source ↗ |
| CyberSecEval | Deprecated | 61.2% | 10 | 2024-05-25 | independent | Source ↗ |
| MMLU-Redux | Saturated | 77.4% | 25 | 2024-05-24 | independent | Source ↗ |
| QuALITY | Nearing Saturation | 69.5% | 78 | 2024-05-24 | independent | Source ↗ |
| SWE-bench Verified | Saturated | 53% | 25 | 2024-05-24 | independent | Source ↗ |
| MGSM | Nearing Saturation | 69.4% | 65 | 2024-05-23 | independent | Source ↗ |
| LongBench | Nearing Saturation | 67.9% | 65 | 2024-05-23 | independent | Source ↗ |
| SWE-bench | Deprecated | 46.2% | 10 | 2024-05-23 | independent | Source ↗ |
| AGIEval | Saturated | 66.2% | 25 | 2024-05-22 | independent | Source ↗ |
| QASPER | Nearing Saturation | 65.6% | 65 | 2024-05-22 | independent | Source ↗ |
| TruthfulQA | Saturated | 69.5% | 25 | 2024-05-22 | independent | Source ↗ |
| SuperGLUE | Saturated | 78.5% | 25 | 2024-05-21 | independent | Source ↗ |
| GSM-Hard | Deprecated | 70.1% | 10 | 2024-05-21 | independent | Source ↗ |
| MT-Bench | Nearing Saturation | 7.5% | 75 | 2024-05-20 | independent | Source ↗ |
| BIG-Bench Hard | Saturated | 75.5% | 25 | 2024-05-20 | independent | Source ↗ |
| CodeXGLUE | Deprecated | 63.9% | 23 | 2024-05-20 | independent | Source ↗ |
| MBPP+ | Nearing Saturation | 69.4% | 65 | 2024-05-19 | independent | Source ↗ |
| WinoGrande | Saturated | 80.1% | 28 | 2024-05-19 | independent | Source ↗ |
| IFEval | Deprecated | 71.2% | 25 | 2024-05-19 | independent | Source ↗ |
| GPQA | Nearing Saturation | 69.5% | 78 | 2024-05-19 | independent | Source ↗ |
| HumanEval+ | Saturated | 73.4% | 25 | 2024-05-18 | independent | Source ↗ |
| Needle-in-a-Haystack | Saturated | 95.8% | 44 | 2024-05-18 | independent | Source ↗ |
| CommonsenseQA | Saturated | 80.9% | 28 | 2024-05-18 | independent | Source ↗ |
| GSM8K | Saturated | 78.8% | 25 | 2024-05-17 | independent | Source ↗ |
| GLUE | Saturated | 83.2% | 28 | 2024-05-17 | independent | Source ↗ |
| MBPP | Saturated | 72.1% | 25 | 2024-05-17 | independent | Source ↗ |
| HumanEval | Saturated | 77.5% | 25 | 2024-05-16 | independent | Source ↗ |
| MMLU | Saturated | 80.6% | 25 | 2024-05-16 | independent | Source ↗ |
| ARC-AGI | Saturated | 72.4% | 25 | 2024-05-16 | independent | Source ↗ |
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run hanooman-7b
# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "hanooman-7b",
"messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
}'Evaluation & Sourcing Notes
Scores listed for Hanooman 7B 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.