Model comparison
Gemini 3.5 Flash Lite vs GPT-6 Astra
GPT-6 Astra is the stronger model overall, scoring 70.3 to 41.8 on the Noometry Index. Gemini 3.5 Flash Lite costs 24× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.
Last verified . 34 shared benchmarks.
GoogleSummary
- They share 34 benchmarks with published results for both. Gemini 3.5 Flash Lite scores higher in 0 categories and GPT-6 Astra in 10 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Astra leads 93.4 to 25.7.
- The biggest single-benchmark swing is FrontierMath Tier 4: 0% for Gemini 3.5 Flash Lite and 97.6% for GPT-6 Astra.
- Gemini 3.5 Flash Lite is cheaper at $0.30 / $2.50 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 1.05M.
Side by side
| Gemini 3.5 Flash Lite | GPT-6 Astra | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 41.8 | 70.3 |
| Released | 2026-07-21 | 2026-09-03 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.30 | $10 |
| Output $ / M tokens | $2.50 | $50 |
| Results tracked | 38 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
Gemini 3.5 Flash Lite: 42.1 (#96), GPT-6 Astra: 70.5 (#3)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| LMArena WebDev | 1440 | 1792 |
| SciCode | 41.3% | 56.5% |
| WeirdML | 39% | 93.6% |
| LMArena Coding | 1455 | 1488 |
| ALE-Bench | 765.27 | 2,951 |
| DeepSWE | — | 74.1% |
| FrontierCode | — | 53.3% |
| FrontierSWE | — | 65.5% |
| MirrorCode | — | 46.7% |
Agentic & Tool Use GPT-6 Astra leads
Gemini 3.5 Flash Lite: 28.6 (#82), GPT-6 Astra: 52.3 (#3)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| APEX-Agents | — | 64.7% |
| Remote Labor Index | — | 20.8% |
| BALROG | — | 68.3% |
| GDP.pdf | 10% | — |
| Vending-Bench 2 | — | 15,515 |
Reasoning GPT-6 Astra leads
Gemini 3.5 Flash Lite: 27.8 (#95), GPT-6 Astra: 85.1 (#1)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| ARC-AGI-2 | 10.3% | 95% |
| NYT Connections (extended) | 60.4% | 98.1% |
| ARC-AGI-1 | 53.5% | 98.5% |
| CritPt | 0% | 31.7% |
| Chess Puzzles | 22% | 72% |
| LMArena Hard Prompts | 1441 | 1464 |
| Mystery Game Puzzles | 19% | 84% |
| DTBench | 83.5% | 97.3% |
| LMCA | 37.6% | 64.4% |
| Epoch Capabilities Index | 145.12 | 166.6 |
| EBR-Bench | — | 76.2% |
| Bench to the Future 3 | — | 0.14 |
Math GPT-6 Astra leads
Gemini 3.5 Flash Lite: 25.7 (#245), GPT-6 Astra: 93.4 (#1)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| FrontierMath (Tiers 1-3) | 26% | 93.7% |
| FrontierMath Tier 4 | 0% | 97.6% |
| OTIS Mock AIME 2024-2025 | 71.1% | 100% |
| ProofBench | 13% | 99% |
| LMArena Math | 1431 | 1463 |
| FrontierMath Erdős | — | 2.9% |
Knowledge GPT-6 Astra leads
Gemini 3.5 Flash Lite: 50.0 (#66), GPT-6 Astra: 75.3 (#1)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| GPQA Diamond | 83.3% | 95.8% |
| LMArena Expert | 1430 | 1486 |
| Humanity's Last Exam | — | 54.8% |
| SimpleQA Verified | — | 75.6% |
| Vectara Hallucination Rate | — | 8.7% |
Multimodal GPT-6 Astra leads
Gemini 3.5 Flash Lite: 40.3 (#38), GPT-6 Astra: 55.1 (#3)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| LMArena Vision | 1266 | 1280 |
| Blueprint-Bench 2 | — | 49.7% |
| Furniture Assembly | — | 80% |
| LMArena Document | — | 1468 |
Multilingual Too close to call
Gemini 3.5 Flash Lite: 53.8 (#59), GPT-6 Astra: 54.0 (#53)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1428 | 1431 |
| LMArena Chinese | 1467 | 1495 |
| LMArena French | 1452 | 1467 |
| LMArena Russian | 1443 | 1434 |
| LMArena Spanish | 1440 | 1412 |
| LMArena German | 1454 | — |
| LMArena Japanese | 1428 | — |
| LMArena Korean | 1406 | — |
Instruction Following GPT-6 Astra leads
Gemini 3.5 Flash Lite: 75.0 (#77), GPT-6 Astra: 76.6 (#37)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1420 | 1454 |
Long Context Too close to call
Gemini 3.5 Flash Lite: 43.8 (#80), GPT-6 Astra: 44.7 (#57)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1435 | 1460 |
Writing & Preference GPT-6 Astra leads
Gemini 3.5 Flash Lite: 64.4 (#55), GPT-6 Astra: 75.7 (#7)
| Benchmark | Gemini 3.5 Flash Lite | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1436 | 1443 |
| LMArena Creative Writing | 1419 | 1423 |
| EQ-Bench Creative Writing | 1559 | 2173 |
| LMArena Multi-Turn | 1445 | 1457 |
Frequently asked questions
Is Gemini 3.5 Flash Lite better than GPT-6 Astra?
GPT-6 Astra is the stronger model overall, scoring 70.3 to 41.8 on the Noometry Index. Gemini 3.5 Flash Lite costs 24× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.
Which is cheaper, Gemini 3.5 Flash Lite or GPT-6 Astra?
Gemini 3.5 Flash Lite is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; GPT-6 Astra lists at $10 and $50.
Is Gemini 3.5 Flash Lite or GPT-6 Astra better for coding?
GPT-6 Astra scores higher on coding benchmarks: 70.5 versus 42.1 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Astra does, with 1.05M tokens against 1.05M.
How many benchmarks do Gemini 3.5 Flash Lite and GPT-6 Astra share?
34 benchmarks have published results for both models. Gemini 3.5 Flash Lite has 38 scored results on Noometry and GPT-6 Astra has 51.