Grok 4.20 (Non-Reasoning)
Grok model for agentic tool use, reasoning, coding, and live assistance
Shogo Agent Index
Estimated#49 of 9243.9range 33.3–54.5 · 19% of weight measured
Agentic coding
45
23% measured
Long-horizon autonomy
59
30% measured
Knowledge-work agents
34
estimated
Computer use
32
estimated
Security & context learning
45
50% measured
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 75 | |
| Terminal-Bench 2.0 | 57.3% | 8% | Grok CLI | |
| DeepSWE | est. | 6% | predicted 42 | |
| FrontierSWE | est. | 5% | predicted 11 | |
| CursorBench | est. | 4% | predicted 34 | |
| GSO-Bench | est. | 4% | predicted 21 | |
| METR Time Horizons | est. | 10% | predicted 72 | |
| Vending-Bench 2 | $4,663 | 6% | source | |
| PostTrainBench | est. | 4% | predicted 23 | |
| APEX-Agents | est. | 8% | predicted 39 | |
| GDPval | est. | 7% | predicted 42 | |
| Remote Labor Index | est. | 5% | predicted 3 | |
| DeepResearch Bench | est. | 5% | predicted 47 | |
| OSWorld 2.0 | est. | 6% | predicted 7 | |
| OSWorld | est. | 4% | predicted 69 | |
| Cybench | est. | 5% | predicted 68 | |
| CL-bench | 22.2% | 5% | source |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| ARC-AGI | Reasoning | 89.5% | source |
| ARC-AGI-2 | Reasoning | 65.1% | source |
| Chess Puzzles | Reasoning | 29.0% | source |
| DTBench | Knowledge | 90.1% | source |
| FrontierMath Tiers 1-3 | Math | 44.9% | source |
| FrontierMath Tier 4 | Math | 17.1% | source |
| GPQA Diamond | Reasoning | 89.3% | source |
| LMCA | Knowledge | 38.7% | source |
| OTIS Mock AIME | Math | 92.2% | source |
| ProofBench | Math | 14.0% | source |
| SimpleQA Verified | Knowledge | 30.2% | source |
| WebDev Arena | Coding | 1374 | source |