Model comparison
Claude Sonnet 4.6 vs Gemini 3.8 Flash
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 50.3 on the Noometry Index.
Last verified . 42 shared benchmarks.
GoogleSummary
- They share 42 benchmarks with published results for both. Claude Sonnet 4.6 scores higher in 0 categories and Gemini 3.8 Flash in 10 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 46.1.
- The biggest single-benchmark swing is Chess Puzzles: 13% for Claude Sonnet 4.6 and 61% for Gemini 3.8 Flash.
- Gemini 3.8 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.6.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Sonnet 4.6 | Gemini 3.8 Flash | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 50.3 | 61.8 |
| Released | 2026-02-17 | 2026-09-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 66K |
| Input $ / M tokens | $3 | $0.75 |
| Output $ / M tokens | $15 | $3.75 |
| Results tracked | 57 | 50 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.8 Flash leads
Claude Sonnet 4.6: 46.3 (#67), Gemini 3.8 Flash: 59.2 (#15)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| DeepSWE | 29.9% | 73.8% |
| FrontierCode | 24.3% | 41.2% |
| LMArena WebDev | 1522 | 1584 |
| SciCode | 46.8% | 56.6% |
| WeirdML | 66.1% | 84.8% |
| LMArena Coding | 1504 | 1510 |
| ALE-Bench | 1,327 | 1,270 |
| SWE-bench Verified | 75.2% | — |
| CursorBench | — | 39.6% |
| FrontierSWE | — | 19.6% |
Agentic & Tool Use Gemini 3.8 Flash leads
Claude Sonnet 4.6: 39.1 (#28), Gemini 3.8 Flash: 41.8 (#21)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| APEX-Agents | 43% | 64.3% |
| GDP.pdf | 18% | 23.4% |
| Vending-Bench 2 | 7,204 | 5,094 |
| Terminal-Bench | 53.4% | — |
| OSWorld 2.0 | 9.3% | — |
| Remote Labor Index | — | 5.8% |
| DeepResearch Bench | 54.9% | — |
| OSWorld | 72.1% | — |
| ExploitBench | 23.6% | — |
| GBAEval | 48.8% | — |
| LMArena Search | 1221 | — |
Reasoning Gemini 3.8 Flash leads
Claude Sonnet 4.6: 46.1 (#45), Gemini 3.8 Flash: 76.9 (#5)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| ARC-AGI-2 | 60.4% | 89.2% |
| NYT Connections (extended) | 80.9% | 97.4% |
| ARC-AGI-1 | 86.5% | 98.5% |
| CritPt | 3.1% | 18.3% |
| Chess Puzzles | 13% | 61% |
| LMArena Hard Prompts | 1484 | 1508 |
| Mystery Game Puzzles | 16% | 47% |
| DTBench | 89.9% | 95.7% |
| LMCA | 46.5% | 52.9% |
| Epoch Capabilities Index | 152.24 | 156.71 |
| Thematic Generalization | 76.3% | — |
| Surface Evolver Bench | — | 76.9% |
| ForecastBench | 62 | — |
Math Gemini 3.8 Flash leads
Claude Sonnet 4.6: 52.9 (#49), Gemini 3.8 Flash: 65.3 (#28)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 85.8% | 98.9% |
| ProofBench | 45% | 48% |
| LMArena Math | 1462 | 1528 |
| FrontierMath (Tiers 1-3) | — | 68.4% |
| FrontierMath Tier 4 | — | 22% |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 8.3% | — |
Knowledge Gemini 3.8 Flash leads
Claude Sonnet 4.6: 51.7 (#65), Gemini 3.8 Flash: 74.8 (#2)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| GPQA Diamond | 87.4% | 95.4% |
| SimpleQA Verified | 35.5% | 69.7% |
| LMArena Expert | 1500 | 1524 |
| Humanity's Last Exam | — | 44.5% |
| Vectara Hallucination Rate | 10.6% | — |
Multimodal Gemini 3.8 Flash leads
Claude Sonnet 4.6: 38.0 (#68), Gemini 3.8 Flash: 40.7 (#45)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Vision | 1283 | 1314 |
| Blueprint-Bench 2 | 6.7% | 38.6% |
| Furniture Assembly | — | 31.7% |
| LMArena Document | 1482 | — |
Multilingual Gemini 3.8 Flash leads
Claude Sonnet 4.6: 54.4 (#41), Gemini 3.8 Flash: 58.0 (#5)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Non-English | 1440 | 1491 |
| LMArena Chinese | 1491 | 1554 |
| LMArena French | 1465 | 1498 |
| LMArena German | 1428 | 1493 |
| LMArena Japanese | 1420 | 1502 |
| LMArena Korean | 1411 | 1459 |
| LMArena Russian | 1440 | 1515 |
| LMArena Spanish | 1464 | 1485 |
Instruction Following Too close to call
Claude Sonnet 4.6: 77.4 (#25), Gemini 3.8 Flash: 78.0 (#13)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Instruction Following | 1475 | 1490 |
Long Context Too close to call
Claude Sonnet 4.6: 45.3 (#44), Gemini 3.8 Flash: 46.3 (#24)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Longer Query | 1479 | 1508 |
Writing & Preference Gemini 3.8 Flash leads
Claude Sonnet 4.6: 70.2 (#22), Gemini 3.8 Flash: 72.2 (#15)
| Benchmark | Claude Sonnet 4.6 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Text | 1458 | 1499 |
| LMArena Creative Writing | 1435 | 1492 |
| EQ-Bench Creative Writing | 1810 | 1748 |
| LMArena Multi-Turn | 1464 | 1501 |
| EQ-Bench 4 | 1207 | — |
Frequently asked questions
Is Claude Sonnet 4.6 better than Gemini 3.8 Flash?
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 50.3 on the Noometry Index.
Which is cheaper, Claude Sonnet 4.6 or Gemini 3.8 Flash?
Gemini 3.8 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; Claude Sonnet 4.6 lists at $3 and $15.
Is Claude Sonnet 4.6 or Gemini 3.8 Flash better for coding?
Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 46.3 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.8 Flash does, with 1.05M tokens against 1M.
How many benchmarks do Claude Sonnet 4.6 and Gemini 3.8 Flash share?
42 benchmarks have published results for both models. Claude Sonnet 4.6 has 57 scored results on Noometry and Gemini 3.8 Flash has 50.