Gemini 3.8 Flash vs Gemini 4 Argon
Gemini 4 Argon leads Gemini 3.8 Flash by 13.3 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Gemini 3.8 Flash wins 0 and Gemini 4 Argon wins 22 of 23 shared benchmarks. Gemini 3.8 Flash is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | Gemini 3.8 Flash | Gemini 4 Argon |
|---|---|---|
| Agentic coding | 63 | 77 |
| Long-horizon autonomy | 64 | 78 |
| Knowledge-work agents | 47 | 59 |
| Computer use | 44 | 61 |
| Security & context learning | 53 | 61 |
Shared benchmarks
| Benchmark | Gemini 3.8 Flash | Gemini 4 Argon |
|---|---|---|
| APEX-Agentsindex | 64.3% | 82.4% |
| Vending-Bench 2index | $5,094 | $13,718 |
| FrontierSWEindex | 19.6% | 55.0% |
| Remote Labor Indexindex | 5.8% | 17.1% |
| CursorBenchindex | 39.6% | 52.9% |
| ARC-AGI | 98.5% | 98.5% |
| ARC-AGI-2 | 89.2% | 91.3% |
| Chess Puzzles | 58.0% | 59.0% |
| CritPt | 18.3% | 27.1% |
| DTBench | 95.7% | 97.6% |
| FrontierCode | 41.2% | 47.5% |
| FrontierMath Tier 4 | 22.0% | 78.0% |
| FrontierMath Tiers 1-3 | 68.4% | 86.0% |
| GPQA Diamond | 95.4% | 96.4% |
| LMCA | 52.9% | 63.7% |
| MirrorCode | 38.9% | 58.9% |
| OTIS Mock AIME | 98.9% | 100.0% |
| ProofBench | 48.0% | 99.0% |
| SciCode | 56.6% | 61.8% |
| SimpleBench | 82.4% | 91.7% |
| SimpleQA Verified | 69.7% | 76.0% |
| WebDev Arena | 1568 | 1680 |
| WeirdML v3 | 7.4% | 20.3% |
Specs and price
| Gemini 3.8 Flash | Gemini 4 Argon | |
|---|---|---|
| Agent Index | 56.2 (#21) | 69.5 (#6) |
| Agent $/M | $0.35 | $0.84 |
| Input / output $/M | $0.75 / $3.75 | $2.00 / $10.00 |
| Context | 1M | 1M |
| Released | Sep 2, 2026 | Oct 13, 2026 |
| Weights | Closed | Closed |
| API hosts | 6 | 5 |
Don’t pick one. Route.
Shogo can run Gemini 3.8 Flash and Gemini 4 Argon side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo