DeepSeek V4 Flash
Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work
Shogo Agent Index
UnrankedNot ranked yet: 0 agentic benchmarks measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.
Agentic coding
35
estimated
Long-horizon autonomy
42
estimated
Knowledge-work agents
27
estimated
Computer use
22
estimated
Security & context learning
35
estimated
Price
- Input
- $0.14/M
- Output
- $0.28/M
- Cache read
- $0.028/M
- 3:1 blend
- $0.18/M
- Agent-effective
- $0.060/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 68 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 40 | |
| DeepSWE | est. | 6% | predicted 28 | |
| FrontierSWE | est. | 5% | predicted 5 | |
| CursorBench | est. | 4% | predicted 28 | |
| GSO-Bench | est. | 4% | predicted 10 | |
| METR Time Horizons | est. | 10% | predicted 66 | |
| Vending-Bench 2 | est. | 6% | predicted 17 | |
| PostTrainBench | est. | 4% | predicted 18 | |
| APEX-Agents | est. | 8% | predicted 27 | |
| GDPval | est. | 7% | predicted 33 | |
| Remote Labor Index | est. | 5% | predicted 1 | |
| DeepResearch Bench | est. | 5% | predicted 45 | |
| OSWorld 2.0 | est. | 6% | predicted 3 | |
| OSWorld | est. | 4% | predicted 50 | |
| Cybench | est. | 5% | predicted 52 | |
| CL-bench | est. | 5% | predicted 18 |