Model comparison
Claude Sonnet 4 vs Gemini 2.5 Flash
Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 39.0 on the Noometry Index. Gemini 2.5 Flash costs 7.1× less per token, which makes it the better buy when Claude Sonnet 4's lead doesn't matter for your workload.
Last verified . 49 shared benchmarks.
GoogleSummary
- They share 49 benchmarks with published results for both. Claude Sonnet 4 scores higher in 6 categories and Gemini 2.5 Flash in 4 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Gemini 2.5 Flash leads 41.2 to 25.6.
- The biggest single-benchmark swing is GeoBench: 37% for Claude Sonnet 4 and 76% for Gemini 2.5 Flash.
- Gemini 2.5 Flash is cheaper at $0.30 / $2.50 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- Gemini 2.5 Flash accepts more context: 1.05M tokens versus 200K.
Side by side
| Claude Sonnet 4 | Gemini 2.5 Flash | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 40.6 | 39.0 |
| Released | 2025-05-22 | 2025-04-17 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1.05M |
| Max output | 64K | 66K |
| Input $ / M tokens | $3 | $0.30 |
| Output $ / M tokens | $15 | $2.50 |
| Results tracked | 58 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4 leads
Claude Sonnet 4: 44.0 (#70), Gemini 2.5 Flash: 35.5 (#210)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| SWE-bench Verified (bash only) | 64.9% | 28.7% |
| Aider Polyglot | 61.3% | 55.1% |
| WeirdML | 46.1% | 41.9% |
| LMArena Coding | 1414 | 1424 |
| ALE-Bench | 655.35 | 661.88 |
| SciCode | 40% | — |
| GSO | 4.9% | — |
Agentic & Tool Use Claude Sonnet 4 leads
Claude Sonnet 4: 38.3 (#29), Gemini 2.5 Flash: 30.6 (#67)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| TheAgentCompany | 33.1% | 41.1% |
| Terminal-Bench | — | 17.1% |
| Berkeley Function Calling Leaderboard | — | 56.2% |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| BALROG | — | 33.5% |
| METR Time Horizons | 62% | — |
| Vending-Bench 2 | — | 548.84 |
Reasoning Claude Sonnet 4 leads
Claude Sonnet 4: 23.0 (#169), Gemini 2.5 Flash: 18.2 (#271)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| ARC-AGI-2 | 5.9% | 2.5% |
| SimpleBench | 45.5% | 41.2% |
| Kagi LLM Benchmark | 73% | 56.8% |
| ARC-AGI-1 | 40% | 33.3% |
| CritPt | 0.3% | 1.1% |
| EnigmaEval | 3.1% | 2.7% |
| LMArena Hard Prompts | 1372 | 1422 |
| DTBench | 77.1% | 76.5% |
| LMCA | 29% | 27.5% |
| Epoch Capabilities Index | 141.69 | 143.04 |
| ForecastBench | 60.2 | 60.6 |
Math Claude Sonnet 4 leads
Claude Sonnet 4: 41.5 (#71), Gemini 2.5 Flash: 38.1 (#87)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 73.1% |
| Omni-MATH | 60.2% | 38.5% |
| LMArena Math | 1375 | 1415 |
| FrontierMath (Feb 2025 set) | 4.1% | 4.8% |
| FrontierMath Tier 4 (v1) | 0% | 4.2% |
| MATH Level 5 | 84.4% | — |
Knowledge Claude Sonnet 4 leads
Claude Sonnet 4: 41.6 (#99), Gemini 2.5 Flash: 36.1 (#162)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| Humanity's Last Exam | 7.8% | 12.1% |
| MMLU-Pro | 84.3% | 63.9% |
| Confabulations | 13.2% | 16.8% |
| Vectara Hallucination Rate | 10.3% | 7.8% |
| GPQA (HELM) | 70.6% | 39% |
| LMArena Expert | 1371 | 1426 |
| GPQA Diamond | 79.2% | — |
Multimodal Gemini 2.5 Flash leads
Claude Sonnet 4: 25.6 (#116), Gemini 2.5 Flash: 41.2 (#31)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Vision | 1191 | 1254 |
| GeoBench | 37% | 76% |
| VPCT | 34% | 46.2% |
| MindCube | 44.8% | — |
| SpatialViz-Bench | — | 36.9% |
Multilingual Gemini 2.5 Flash leads
Claude Sonnet 4: 46.9 (#151), Gemini 2.5 Flash: 52.5 (#83)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Non-English | 1333 | 1410 |
| LMArena Chinese | 1351 | 1450 |
| LMArena French | 1364 | 1433 |
| LMArena German | 1331 | 1419 |
| LMArena Japanese | 1301 | 1403 |
| LMArena Korean | 1293 | 1385 |
| LMArena Russian | 1355 | 1415 |
| LMArena Spanish | 1356 | 1421 |
Instruction Following Gemini 2.5 Flash leads
Claude Sonnet 4: 71.8 (#141), Gemini 2.5 Flash: 75.9 (#50)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| IFEval | 84% | 89.8% |
| LMArena Instruction Following | 1375 | 1405 |
Long Context Gemini 2.5 Flash leads
Claude Sonnet 4: 33.6 (#254), Gemini 2.5 Flash: 47.5 (#16)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| Fiction.LiveBench | 46.9% | 77.8% |
| LMArena Longer Query | 1398 | 1419 |
Writing & Preference Claude Sonnet 4 leads
Claude Sonnet 4: 57.3 (#127), Gemini 2.5 Flash: 54.0 (#153)
| Benchmark | Claude Sonnet 4 | Gemini 2.5 Flash |
|---|---|---|
| LMArena Text | 1351 | 1417 |
| LMArena Creative Writing | 1345 | 1400 |
| Short-Story Creative Writing | 81.4% | 76.5% |
| EQ-Bench Creative Writing | 1483 | 1137 |
| WildBench | 83.8% | 81.7% |
| LMArena Multi-Turn | 1376 | 1408 |
Frequently asked questions
Is Claude Sonnet 4 better than Gemini 2.5 Flash?
Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 39.0 on the Noometry Index. Gemini 2.5 Flash costs 7.1× less per token, which makes it the better buy when Claude Sonnet 4's lead doesn't matter for your workload.
Which is cheaper, Claude Sonnet 4 or Gemini 2.5 Flash?
Gemini 2.5 Flash is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; Claude Sonnet 4 lists at $3 and $15.
Is Claude Sonnet 4 or Gemini 2.5 Flash better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 44.0 versus 35.5 in the Noometry coding category.
Which has the bigger context window?
Gemini 2.5 Flash does, with 1.05M tokens against 200K.
How many benchmarks do Claude Sonnet 4 and Gemini 2.5 Flash share?
49 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Gemini 2.5 Flash has 54.