Gemini 3.5 Flash vs GPT-6 Sol
Full profiles: Gemini 3.5 Flash (Google) and GPT-6 Sol (OpenAI): every benchmark, price and rank.
Updated · Data from 12 sources
As of 9 Nov 2026, Gemini 3.5 Flash and GPT-6 Sol are neck and neck on quality (62.5 vs 62.9 on the BenchLeader Index, inside the ±4.3 margin), and Gemini 3.5 Flash is 1.2× cheaper at $3.38 vs $4.00 per 1M tokens blended. Choose Gemini 3.5 Flash for multimodal (66 vs 61), agents & tools (63 vs 60), and human preference (67 vs 64) and fast first answers (12.4 s vs 72.1 s), where it also streams faster; choose GPT-6 Sol for maths (69 vs 58), coding (67 vs 60), and reasoning (66 vs 61).
Key facts (as of 9 Nov 2026)
- Gemini 3.5 Flash scores 62.5 on the BenchLeader Index and GPT-6 Sol scores 62.9 (BenchLeader, 9 Nov 2026).
- Gemini 3.5 Flash costs $3.38 per 1M tokens blended (list price, 9 Nov 2026).
- GPT-6 Sol costs $4.00 per 1M tokens blended (list price, 9 Nov 2026).
- Gemini 3.5 Flash streams faster than GPT-6 Sol (Artificial Analysis, 7-day median, 9 Nov 2026).
- Of the 56 benchmarks both models have, GPT-6 Sol leads 40 (BenchLeader, 9 Nov 2026).
- On Arena WebDev, Gemini 3.5 Flash is rated 1500 and GPT-6 Sol 1674 (LMArena, as published by 9 Nov 2026).
Gemini 3.5 Flash benchmarks 
GPT-6 Sol benchmarks Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | GPT-6 Sol | 67 vs 60 category score; Arena WebDev 1674 vs 1500 | Medium |
| Agents & tool use | Gemini 3.5 Flash | 63 vs 60 category score; Terminal-Bench 2.1 (Vals) 74.2% vs 83.2% | Medium |
| Maths | GPT-6 Sol | 69 vs 58 category score; FrontierMath Tiers 1–3 89.8% vs 62.8% | High |
| Reasoning | GPT-6 Sol | 66 vs 61 category score; GPQA Diamond 94.3% vs 92.8% | Medium |
| Knowledge | Gemini 3.5 Flash | 64 vs 62 category score; SimpleQA Verified 66.2% vs 60.7% | Medium |
| Long documents | GPT-6 Sol | 67 vs 61 category score; AA-LCR v1.1 83.7% vs 73.3% | Medium |
| Images and other media | Gemini 3.5 Flash | 66 vs 61 category score; Arena Vision 1308 vs 1232 | Medium |
| Chat quality (human preference) | Gemini 3.5 Flash | 67 vs 64 on Arena votes | Medium |
| Lowest cost | Gemini 3.5 Flash | $3.38 vs $4.00 per 1M blended; agentic coding session $0.126 vs $0.160 | – |
| Fastest first answer | Gemini 3.5 Flash | 12.4 s (Gemini 3.5 Flash (high)) vs 72.1 s (GPT-6 Sol (max)) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Gemini 3.5 Flash (high) and GPT-6 Sol (max) are neck and neck on quality (62.5 vs 62.9, within the ±4.3 margin).
- Gemini 3.5 Flash (high) is stronger in agents & tools, human preference, instruction following, knowledge, multimodal.
- GPT-6 Sol (max) is stronger in coding, long context, maths, reasoning.
- Gemini 3.5 Flash (high) is 1.2× cheaper ($3.38 vs $4.00 per 1M blended).
- Gemini 3.5 Flash (high) streams 1.5× faster (181 vs 121 tokens per second).
At a glance
| Spec | Gemini 3.5 Flash | GPT-6 Sol |
|---|---|---|
| Developer | OpenAI | |
| Released | 19 May 2026 | 22 Sep 2026 |
| Context window (tokens) | 1,048,576 | 1,050,000 |
| Compared at | Gemini 3.5 Flash (high) | GPT-6 Sol (max) |
| Open weights | No | No |
| Price in / out / cached input per 1M | $1.50 / $9.00 / $0.150 | $2.00 / $10.00 / $0.200 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | Gemini 3.5 Flash (high) | GPT-6 Sol (max) | Leader |
|---|---|---|---|
| Reasoning | 60.7 | 66.2 | GPT-6 Sol |
| Coding | 59.7 | 66.5 | GPT-6 Sol |
| Agents & tools | 62.9 | 59.6 | Gemini 3.5 Flash |
| Maths | 58.1 | 68.5 | GPT-6 Sol |
| Knowledge | 63.8 | 61.5 | Gemini 3.5 Flash |
| Instruction following | 74.4 | – | not comparable |
| Human preference | 66.9 | 64.2 | Gemini 3.5 Flash |
| Multimodal | 66.2 | 60.7 | Gemini 3.5 Flash |
| Long context | 61.2 | 66.6 | GPT-6 Sol |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Not drawn, measured for fewer than two of these models: Instruction following.
ModelIndex
Where sources disagree
Across the 56 benchmarks both have, Gemini 3.5 Flash leads 16, GPT-6 Sol leads 40. Every overall index that rates both puts Gemini 3.5 Flash ahead.
| Overall index | Gemini 3.5 Flash (high) | GPT-6 Sol (max) | Leader |
|---|---|---|---|
| BenchLeader Index | 62.5 | 62.9 | neck and neck |
| Arena Text | 1478 | 1457 | Gemini 3.5 Flash |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | Gemini 3.5 Flash (high) | GPT-6 Sol (max) |
|---|---|---|
| Blended price per 1M tokens | $3.38 | $4.00 |
| Input / output per 1M | $1.50 / $9.00 | $2.00 / $10.00 |
| Cached input per 1M | $0.150 | $0.200 |
| Price per task: chat reply | $0.0033 | $0.0038 |
| Price per task: summarize a 30-page report | $0.023 | $0.030 |
| Price per task: code edit | $0.022 | $0.027 |
| Price per task: agentic coding session | $0.126 | $0.160 |
| Price per task: structured extraction | $0.0048 | $0.0060 |
| Agentic coding session, input mostly cached | $0.065 | $0.079 |
| Cost per benchmark task (AA Index) | – | $2.47 |
| Output tokens per task (AA Index) | – | 54k |
| Of which reasoning | – | 35k |
| Time per task (AA Index) | – | 7 min |
| Output speed | 181 tok/s | 121 tok/s |
| Time to first answer | 12.4 s | 72.1 s |
| Time to a complete answer | 27.4 s | 130.2 s |
Each model is compared at its best-ranked reasoning-effort setting. Gemini 3.5 Flash (high) scores 62.5 and takes 12 s to its first answer token; Gemini 3.5 Flash (medium) scores 61.9 in 10 s, Gemini 3.5 Flash (minimal) scores 55.6 in 3 s. GPT-6 Sol (max) scores 62.9 and takes 72 s to its first answer token; GPT-6 Sol (xhigh) scores 58.6 in 24 s, GPT-6 Sol (high) scores 58.2 in 9 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
135 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | Gemini 3.5 Flash (high) | GPT-6 Sol (max) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 62.5 | 62.9 |
| Agents & tools score | BenchLeader | 62.9 | 59.6 |
| Coding score | BenchLeader | 59.7 | 66.5 |
| Human preference score | BenchLeader | 66.9 | 64.2 |
| Instruction following score | BenchLeader | 74.4 | – |
| Knowledge score | BenchLeader | 63.8 | 61.5 |
| Long context score | BenchLeader | 61.2 | 66.6 |
| Maths score | BenchLeader | 58.1 | 68.5 |
| Multimodal score | BenchLeader | 66.2 | 60.7 |
| Reasoning score | BenchLeader | 60.7 | 66.2 |
| Blended price $/M | BenchLeader | $3.38 | $4.00 |
| Output speed | BenchLeader | 181 tok/s | 121 tok/s |
| Time to first answer | BenchLeader | 12.4 s | 72.1 s |
| Context window | BenchLeader | 1,048,576 tokens | 1,050,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 92.8% | 94.3% |
| SimpleBench | SimpleBench | 76.7%† | – |
| Arena Hard Prompts | Arena (LMArena) | 1496 | 1483 |
| LiveBench 1.0 Reasoning | LiveBench | 82.0% | 88.7% |
| GPQA Diamond (AA) | Artificial Analysis | 92.2% | – |
| Humanity's Last Exam (AA) | Artificial Analysis | 42.7% | 47.9% |
| HLE-Diamond (AA) | Artificial Analysis | – | 45.2% |
| GPQA Diamond (Vals) | Vals AI | 92.7% | – |
| EnigmaEval | Scale AI SEAL | 25.4% | – |
| ARC-AGI-1 | ARC Prize | 92.5% | 95.5% |
| ARC-AGI-2 | ARC Prize | 72.1% | 89.6% |
| ARC-AGI-3 | ARC Prize | – | 23.0% |
| CritPt | Artificial Analysis | 13.1% | 30.9% |
| CritPt v2 hard subset (AA) | Artificial Analysis | – | 66.5% |
| MysteryMechanism | Vals AI | – | 30.2% |
| Chess Puzzles | Epoch AI Benchmarking Hub | 45.0% | 56.0% |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 32.0% | 56.0% |
| LMCA | Epoch AI Benchmarking Hub | 47.1% | 59.1% |
| DTBench | Epoch AI Benchmarking Hub | 94.7% | 97.3% |
| ForecastBench | ForecastBench | 64.4%† | – |
| Coding | |||
| SWE-bench Verified (Epoch) | Epoch AI Benchmarking Hub | 79.3% | – |
| SciCode | SciCode | 53.9% | 57.6% |
| WeirdML | WeirdML | 62.6% | – |
| FrontierCode | Cognition | – | 49.3% |
| Arena Coding | Arena (LMArena) | 1508 | 1520 |
Show all 135 benchmarks
| Metric | Publisher | Gemini 3.5 Flash (high) | GPT-6 Sol (max) |
|---|---|---|---|
| Coding | |||
| Arena WebDev | Arena (LMArena) | 1500 | 1674 |
| LiveBench 1.0 Coding | LiveBench | 78.2% | 81.8% |
| LiveCodeBench | Vals AI | 87.6% | – |
| IOI | Vals AI | – | 82.6% |
| SWE-bench (Vals) | Vals AI | 78.8% | – |
| Code Migration | Vals AI | 26.8% | 57.2% |
| ProgramBench | Vals AI | 0.0% | 2.0% |
| Vibe Code Bench v1.1 | Vals AI | 48.7% | 87.8% |
| Surface Evolver Bench | Epoch AI Benchmarking Hub | 58.1%† | – |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | 36.1% | – |
| ALE-Bench | Epoch AI Benchmarking Hub | 911.0 | 2462 |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 6.6% | 43.9% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 78.7% | – |
| GameBench 2 Overall | DuelLab GameBench 2 | 49.4† | 72.9† |
| GameBench 2 by reasoning setting | DuelLab GameBench 2 | 62.6 | 79.3† |
| Design Arena Website | Design Arena | 1266† | 1285† |
| FrontierCode 2.0 (AA) | Artificial Analysis | – | 35.6% |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | – | 50.3% |
| APEX-Agents | Mercor | 27.5%† | 54.3% |
| Arena Agent | Arena (LMArena) | – | 7 |
| LiveBench Agentic Coding | LiveBench | 49.0% | 52.9% |
| Terminal-Bench Hard | Artificial Analysis | 40.9% | – |
| τ²-Bench Telecom (AA) | Artificial Analysis | 95.3% | – |
| Terminal-Bench 2.1 (Vals) | Vals AI | 74.2% | 83.2% |
| MCP Atlas | Scale AI SEAL | 83.6% | – |
| HiL-Bench | Scale AI SEAL | 27.7%† | – |
| τ³-Banking (AA) | Artificial Analysis | 32.2% | – |
| ITBench SRE (AA) | Artificial Analysis | 40.4% | 49.4% |
| Analyst Agent (AA) | Artificial Analysis | 45.0% | – |
| APEX-Agents (AA) | Artificial Analysis | 47.0% | – |
| CyberBench | Vals AI | – | 78.0% |
| Finance Agent v2 | Vals AI | 57.9% | 49.0% |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 2.5% | 1.7% |
| Legal Research Bench | Vals AI | 30.8% | 28.9% |
| SkillsBench | Vals AI | 52.7% | – |
| SREBench | Vals AI | – | 30.5% |
| Tax Agent Bench | Vals AI | 21.6% | 15.0% |
| Terminal-Bench 2.0 (Vals) | Vals AI | 67.4% | – |
| Terminal-Bench 4.0 (Vals) | Vals AI | 6.1% | 44.4% |
| Terminal-Bench Science | Vals AI | 5.7% | 30.0% |
| Web Search Index | Vals AI | 37.0% | – |
| Arena Search | Arena (LMArena) | – | 1229 |
| GBAEval | Epoch AI Benchmarking Hub | 6.7%† | – |
| Vending-Bench 2 | Andon Labs | 5396.4† | 14427.9† |
| AutomationBench | Artificial Analysis | 42.1% | 61.6% |
| GDP.pdf (AA) | Artificial Analysis | 19.8% | 25.2% |
| Harvey LAB-AA v1.1 | Harvey | – | 3.6% |
| EnterpriseOps-Gym | Artificial Analysis | 50.1% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 876 | 1476 |
| Design Arena Agentic | Design Arena | 1168† | 1212† |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | – | 30.0% |
| CWE-Bench (AA) | Artificial Analysis | – | 64.2% |
| DeepSec-Bench (AA) | Artificial Analysis | – | 46.3% |
| CyberGym end-to-end (AA) | Artificial Analysis | – | 0.0% |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 62.8% | 89.8% |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 26.8% | 90.0% |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 95.6% | 100.0% |
| ProofBench | Vals AI | 31.0% | 83.0%† |
| LiveBench Mathematics | LiveBench | 88.2% | 96.4% |
| AIME 2026 | MathArena | 95.0%† | – |
| HMMT February 2026 | MathArena | 95.5%† | – |
| MathArena Apex | MathArena | 32.3%† | – |
| Arena Maths | Arena (LMArena) | 1497 | 1440 |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 66.2% | 60.7% |
| LiveBench 1.0 Data Analysis | LiveBench | 64.9% | 81.2% |
| AA-Omniscience v1.1 | Artificial Analysis | – | 21.4 |
| MMLU-Pro | Vals AI | 89.5% | – |
| LegalBench | Vals AI | 83.6% | – |
| CorpFin | Vals AI | 64.7% | – |
| TaxEval | Vals AI | 74.4% | – |
| PRBench Legal | Scale AI SEAL | – | 38.6%† |
| BioMysteryBench | Vals AI | – | 74.8% |
| Excel Modeling Benchmark | Vals AI | 63.5% | 71.5% |
| MedCode | Vals AI | 55.8% | 47.1% |
| MedScribe | Vals AI | 76.6% | 82.0% |
| Public Benefits Bench v1.1 | Vals AI | 59.5% | 56.6% |
| DrugDiscoveryBench | Scale AI SEAL | 50.0% | – |
| EBR-bench | Epoch AI Benchmarking Hub | 4.8% | 53.3% |
| GDP.pdf | Epoch AI Benchmarking Hub | 14.0%† | 26.4% |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | – | 51.1% |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | – | 39.1% |
| HealthBench Professional (AA) | OpenAI | – | 62.3% |
| Instruction following | |||
| LiveBench Language | LiveBench | 84.6% | 85.3% |
| LiveBench Instruction Following | LiveBench | 75.6% | 68.6% |
| IFBench | Artificial Analysis | 76.3% | – |
| Arena Instruction Following | Arena (LMArena) | 1465 | 1448 |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1478 | 1457 |
| EQ-Bench 4 | EQ-Bench | 1077† | – |
| Arena Creative Writing | Arena (LMArena) | 1465 | 1430 |
| Arena Multi-turn | Arena (LMArena) | 1480 | 1466 |
| Arena Longer Queries | Arena (LMArena) | 1483 | 1466 |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1308 | 1232 |
| MMMU-Pro | Artificial Analysis | 84.3% | 83.0% |
| MMMU-Pro (Vals) | Vals AI | 88.3% | – |
| MortgageTax | Vals AI | 68.1% | – |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 33.6%† | 36.9%† |
| AA-3D | Artificial Analysis | – | 1131 |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | 73.3% | 83.7% |
| Arena Document | Arena (LMArena) | 1455 | – |
| MLCR | Artificial Analysis | 10.0% | 33.9% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 155.2† | 164.1† |
| LiveBench 1.0 | LiveBench | 74.6% | 79.3% |
| AA Intelligence Index v5.0 | Artificial Analysis | – | 44.2 |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 32.6 | 47.6 |
| AA Progress Index v1 | Artificial Analysis | 305.1 | 334 |
| Vals RSI Index | Vals AI | – | 28.1% |
| Vals Science Index | Vals AI | – | 45.0% |
| Vals Multimodal Index | Vals AI | 62.9% | – |
| AA Cyber Index | Artificial Analysis | – | 36.8 |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | – | 13.5%† |
Data as of 2026-11-09. Best configuration of each model; every score links to its source on the model pages.
DuelLab GameBench 2, release gb2-20261103-eaea31df94e6 (https://benchmarks.duellab.org/releases/gb2-20261103-eaea31df94e6/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
What changed
- GPT-6 Sol (max): BenchLeader Index 63.5 → 62.9.
- GPT-6 Sol (max): 2 score revisions.
- GPT-6 Sol (max): 1 score revision.
- GPT-6 Sol (max): 1 score revision.
- Gemini 3.5 Flash (high): 1 score revision.
- GPT-6 Sol (max): 6 new results (AA Intelligence Index v5.0, AA-Omniscience v1.1, CritPt v2 hard subset (AA)…).
- GPT-6 Sol (max): 1 score revision.
- GPT-6 Sol (max): 1 score revision.
- Gemini 3.5 Flash (high): BenchLeader Index 63.2 → 62.6.
- GPT-6 Sol (max): BenchLeader Index 64.1 → 63.5.
Related comparisons
- Gemini 3.5 Flash vs Claude Fable 5.2
- Gemini 3.5 Flash vs Claude Opus 5.5
- Gemini 3.5 Flash vs Gemini 4 Carbon
- GPT-6 Sol vs Claude Fable 5.2
- GPT-6 Sol vs Claude Opus 5.5
- GPT-6 Sol vs Gemini 4 Carbon
- GPT-6 Sol vs GPT-6 Astra
- Gemini 3.1 Pro vs GPT-6 Sol
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Gemini 3.5 Flash vs GPT-6 Sol: questions
- Is Gemini 3.5 Flash better than GPT-6 Sol?
- Gemini 3.5 Flash (high) and GPT-6 Sol (max) are neck and neck on quality (62.5 vs 62.9, within the ±4.3 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 9 Nov 2026, so check the category scores for your use.
- Gemini 3.5 Flash vs GPT-6 Sol: which is cheaper, and will that change?
- Gemini 3.5 Flash is cheaper today: $3.38 against $4.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Gemini 3.5 Flash or GPT-6 Sol better for coding?
- GPT-6 Sol scores higher in coding (67 vs 60 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1500 for Gemini 3.5 Flash against 1674 for GPT-6 Sol.
- Which is faster, Gemini 3.5 Flash or GPT-6 Sol?
- Gemini 3.5 Flash streams faster (181 vs 121 output tokens per second, Artificial Analysis 7-day median).
- Is Gemini 3.5 Flash or GPT-6 Sol better for agents and tool use?
- Gemini 3.5 Flash scores higher in agentic tasks (63 vs 60 on the category index, where 50 is average). On Terminal-Bench 2.1 (Vals), the top shared agents & tools benchmark, it is 74.2% for Gemini 3.5 Flash against 83.2% for GPT-6 Sol.
More questions (4)
- Which is better for long documents?
- GPT-6 Sol scores higher on long-context benchmarks (67 vs 61). Both take about 1.1M tokens: Gemini 3.5 Flash 1,048,576, GPT-6 Sol 1,050,000.
- Which has lower latency?
- Gemini 3.5 Flash answers first: 12.4 s against 72.1 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Do independent benchmarks agree on Gemini 3.5 Flash vs GPT-6 Sol?
- Across the 56 benchmarks both have, Gemini 3.5 Flash leads 16, GPT-6 Sol leads 40. Every overall index that rates both puts Gemini 3.5 Flash ahead.
- When was GPT-6 Sol released?
- GPT-6 Sol was released on 22 Sep 2026.