Claude-Sonnet-3.5
Balanced Claude model for coding, analysis, agent workflows, and cost control
Shogo Agent Index
Estimated#89 of 9214.8range 10.9–18.8 · 19% of weight measured
Agentic coding
14
11% measured
Long-horizon autonomy
25
50% measured
Knowledge-work agents
14
estimated
Computer use
4
estimated
Security & context learning
14
50% measured
Price
- Input
- $2.60/M
- Output
- $13.00/M
- Cache read
- $0.26/M
- 3:1 blend
- $5.20/M
- Agent-effective
- $1.20/M
Agent-effective = 20 input tokens per output token, 85% cache hits. Why.
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 42 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 4 | |
| DeepSWE | est. | 6% | predicted 6 | |
| FrontierSWE | est. | 5% | predicted 0 | |
| CursorBench | est. | 4% | predicted 14 | |
| GSO-Bench | 4.6% | 4% | OpenHands | |
| METR Time Horizons | 45.2% | 10% | source | |
| Vending-Bench 2 | est. | 6% | predicted 1 | |
| PostTrainBench | est. | 4% | predicted 9 | |
| APEX-Agents | est. | 8% | predicted 7 | |
| GDPval | est. | 7% | predicted 12 | |
| Remote Labor Index | est. | 5% | predicted 0 | |
| DeepResearch Bench | est. | 5% | predicted 39 | |
| OSWorld 2.0 | est. | 6% | predicted 0 | |
| OSWorld | est. | 4% | predicted 8 | |
| Cybench | 17.5% | 5% | source | |
| CL-bench | est. | 5% | predicted 11 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| Aider Polyglot | Coding | 51.6% | source |
| BALROG | Long-horizon | 32.6% | source |
| DTBench | Knowledge | 67.8% | source |
| GeoBench | Multimodal | 62.0% | source |
| GPQA Diamond | Reasoning | 55.3% | source |
| Humanity's Last Exam | Reasoning | 4.1% | source |
| MATH Level 5 | Math | 56.9% | source |
| MMLU | Legacy | 87.3% | source |
| OTIS Mock AIME | Math | 8.5% | source |
| SimpleBench | Reasoning | 41.4% | source |
| TheAgentCompany | Knowledge work | 24.0% | source |
| VPCT | Multimodal | 33.0% | source |