Claude Sonnet 5.5 vs Gemini 3.8 Flash
Claude Sonnet 5.5 leads Gemini 3.8 Flash by 12.8 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Sonnet 5.5 wins 19 and Gemini 3.8 Flash wins 4 of 23 shared benchmarks. Gemini 3.8 Flash is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | Claude Sonnet 5.5 | Gemini 3.8 Flash |
|---|---|---|
| Agentic coding | 79 | 63 |
| Long-horizon autonomy | 75 | 64 |
| Knowledge-work agents | 57 | 47 |
| Computer use | 61 | 44 |
| Security & context learning | 61 | 53 |
Shared benchmarks
| Benchmark | Claude Sonnet 5.5 | Gemini 3.8 Flash |
|---|---|---|
| APEX-Agentsindex | 75.5% | 64.3% |
| Vending-Bench 2index | $11,165 | $5,094 |
| FrontierSWEindex | 61.9% | 19.6% |
| CursorBenchindex | 55.5% | 39.6% |
| ARC-AGI | 96.5% | 98.5% |
| ARC-AGI-2 | 83.9% | 89.2% |
| Chess Puzzles | 59.0% | 58.0% |
| CritPt | 31.4% | 18.3% |
| DTBench | 97.3% | 95.7% |
| FrontierCode | 52.1% | 41.2% |
| FrontierMath Tier 4 | 80.5% | 22.0% |
| FrontierMath Tiers 1-3 | 88.8% | 68.4% |
| GPQA Diamond | 95.6% | 95.4% |
| Humanity's Last Exam | 46.1% | 44.5% |
| LMCA | 68.3% | 52.9% |
| MirrorCode | 97.8% | 38.9% |
| OTIS Mock AIME | 100.0% | 98.9% |
| ProofBench | 100.0% | 48.0% |
| SciCode | 61.0% | 56.6% |
| SimpleBench | 75.9% | 82.4% |
| SimpleQA Verified | 46.5% | 69.7% |
| WebDev Arena | 1786 | 1568 |
| WeirdML v3 | 19.9% | 7.4% |
Specs and price
| Claude Sonnet 5.5 | Gemini 3.8 Flash | |
|---|---|---|
| Agent Index | 69 (#9) | 56.2 (#21) |
| Agent $/M | $0.84 | $0.35 |
| Input / output $/M | $2.00 / $10.00 | $0.75 / $3.75 |
| Context | 1M | 1M |
| Released | Sep 28, 2026 | Sep 2, 2026 |
| Weights | Closed | Closed |
| API hosts | 5 | 6 |
Don’t pick one. Route.
Shogo can run Claude Sonnet 5.5 and Gemini 3.8 Flash side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo