Gemini 3.8 Flash vs GPT-5
Full profiles: Gemini 3.8 Flash (Google) and GPT-5 (OpenAI): every benchmark, price and rank.
Updated · Data from 10 sources
As of 16 Nov 2026, Gemini 3.8 Flash and GPT-5 are neck and neck on quality (60.2 vs 59.0 on the BenchLeader Index, inside the ±7.3 margin), and Gemini 3.8 Flash is 2.3× cheaper at $1.50 vs $3.44 per 1M tokens blended. Choose Gemini 3.8 Flash for agents & tools (75 vs 58), reasoning (59 vs 48), and multimodal (67 vs 57) and fast first answers (14.8 s vs 22.9 s), where it also streams faster; choose GPT-5 for coding (57 vs 52) and long context (70 vs 67). Gemini 3.8 Flash's current price runs until 31 Dec 2026, then $3.00.
Key facts (as of 16 Nov 2026)
- Gemini 3.8 Flash scores 60.2 on the BenchLeader Index and GPT-5 scores 59.0 (BenchLeader, 16 Nov 2026).
- Gemini 3.8 Flash costs $1.50 per 1M tokens blended until 31 Dec 2026, then $3.00 (https://ai.google.dev/gemini-api/docs/pricing, checked 20 Oct 2026).
- GPT-5 costs $3.44 per 1M tokens blended (list price, 16 Nov 2026).
- Gemini 3.8 Flash streams faster than GPT-5 (Artificial Analysis, 7-day median, 16 Nov 2026).
- Of the 9 benchmarks both models have, Gemini 3.8 Flash leads 9 (BenchLeader, 16 Nov 2026).
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-5 | 57 vs 52 category score; GameBench 2 by reasoning setting 38.0 vs 59.7 | Medium |
| Agents & tool use | Gemini 3.8 Flash | 75 vs 58 category score | High |
| Reasoning | Gemini 3.8 Flash | 59 vs 48 category score; ARC-AGI-2 82.9% vs 7.5% | High |
| Long documents | GPT-5 | 70 vs 67 category score; AA-LCR v1.1 76.0% vs 84.0% | Medium |
| Images and other media | Gemini 3.8 Flash | 67 vs 57 category score; MMMU-Pro 84.2% vs 74.3% | High |
| Lowest cost | Gemini 3.8 Flash (until 31 Dec 2026) | $1.50 vs $3.44 per 1M blended; agentic coding session $0.060 vs $0.115; from 1 Jan 2027: $3.00 vs $3.44 | – |
| Fastest first answer | Gemini 3.8 Flash | 14.8 s (Gemini 3.8 Flash (medium)) vs 22.9 s (GPT-5 (medium)) 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 3.8 Flash (medium) and GPT-5 (medium) are neck and neck on quality (60.2 vs 59.0, within the ±7.3 margin).
- Gemini 3.8 Flash (medium) is stronger in agents & tools, multimodal, reasoning.
- GPT-5 (medium) is stronger in coding, long context.
- Gemini 3.8 Flash (medium) is 2.3× cheaper ($1.50 vs $3.44 per 1M blended) until 31 Dec 2026, then $3.00 vs $3.44 from 1 Jan 2027.
- Gemini 3.8 Flash (medium) streams 1.4× faster (116 vs 82 tokens per second).
At a glance
| Spec | Gemini 3.8 Flash | GPT-5 |
|---|---|---|
| Developer | OpenAI | |
| Released | 2 Sep 2026 | 7 Aug 2025 |
| Context window (tokens) | 1,048,576 | 400,000 |
| Compared at | Gemini 3.8 Flash (medium) | GPT-5 (medium) |
| Open weights | No | No |
| Price in / out / cached input per 1M | $0.750 / $3.75 / $0.075 | $1.25 / $10.00 / $0.125 |
| Announced price change | $1.50 / $7.50 from 1 Jan 2027 | None announced |
Limits, availability and announced prices from the vendors' own pages: ai.google.dev pricing (checked 20 Oct 2026). Everything else is from the day's data.
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 3.8 Flash (medium) | GPT-5 (medium) | Leader |
|---|---|---|---|
| Reasoning | 58.7 | 47.9 | Gemini 3.8 Flash |
| Coding | 51.6 | 57.4 | GPT-5 |
| Agents & tools | 75.0 | 58.1 | Gemini 3.8 Flash |
| Maths | – | 65.4 | not comparable |
| Knowledge | 57.7 | – | not comparable |
| Instruction following | – | 69.4 | not comparable |
| Multimodal | 66.7 | 56.6 | Gemini 3.8 Flash |
| Long context | 66.7 | 70.4 | GPT-5 |
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: Maths, Knowledge, Instruction following, Human preference.
Where sources disagree
Across the 9 benchmarks both have, Gemini 3.8 Flash leads 9, GPT-5 leads 0. Every overall index that rates both puts Gemini 3.8 Flash ahead.
| Overall index | Gemini 3.8 Flash (medium) | GPT-5 (medium) | Leader |
|---|---|---|---|
| BenchLeader Index | 60.2 | 59.0 | neck and neck |
| Arena Text | 1497† | 1435 | Gemini 3.8 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 3.8 Flash (medium) | GPT-5 (medium) |
|---|---|---|
| Blended price per 1M tokens, until 31 Dec 2026 | $1.50 | $3.44 |
| Input / output per 1M | $0.750 / $3.75 | $1.25 / $10.00 |
| Blended price per 1M from 1 Jan 2027 | $3.00 | $3.44 |
| Input / output per 1M from 1 Jan 2027 | $1.50 / $7.50 | $1.25 / $10.00 |
| Cached input per 1M | $0.075 | $0.125 |
| Price per task: chat reply | $0.0014 | $0.0035 |
| Price per task: summarize a 30-page report | $0.011 | $0.021 |
| Price per task: code edit | $0.010 | $0.022 |
| Price per task: agentic coding session | $0.060 | $0.115 |
| Price per task: structured extraction | $0.0023 | $0.0045 |
| Agentic coding session, input mostly cached | $0.030 | $0.064 |
| Cost per benchmark task (AA Index) | $6.26 | – |
| Output tokens per task (AA Index) | 82k | – |
| Of which reasoning | 47k | – |
| Time per task (AA Index) | 12 min | – |
| Output speed | 116 tok/s | 82 tok/s |
| Time to first answer | 14.8 s | 22.9 s |
| Time to a complete answer, last measured 13 Nov 2026 | 41.8 s | 66.2 s |
Each model is compared at its best-ranked reasoning-effort setting. Gemini 3.8 Flash (medium) scores 60.2 and takes 15 s to its first answer token; Gemini 3.8 Flash (high) scores 59.4 in 20 s, Gemini 3.8 Flash scores 56.4 in 2 s. GPT-5 (medium) scores 59.0 and takes 23 s to its first answer token; GPT-5 (high) scores 58.2 in 49 s, GPT-5 scores 56.6 in 5 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
229 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 3.8 Flash (medium) | GPT-5 (medium) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 60.2 | 59.0 |
| Agents & tools score | BenchLeader | 75.0 | 58.1 |
| Coding score | BenchLeader | 51.6 | 57.4 |
| Knowledge score | BenchLeader | 57.7 | – |
| Long context score | BenchLeader | 66.7 | 70.4 |
| Multimodal score | BenchLeader | 66.7 | 56.6 |
| Reasoning score | BenchLeader | 58.7 | 47.9 |
| Instruction following score | BenchLeader | – | 69.4 |
| Maths score | BenchLeader | – | 65.4 |
| Blended price $/M | BenchLeader | $1.50 | $3.44 |
| Output speed | BenchLeader | 116 tok/s | 82 tok/s |
| Time to first answer | BenchLeader | 14.8 s | 22.9 s |
| Context window | BenchLeader | 1,048,576 tokens | 400,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 95.4%† | 85.3% |
| Humanity's Last Exam | Scale AI / CAIS | 44.5%† | 25.3%† |
| SimpleBench | SimpleBench | 82.4%† | 56.7%† |
| Arena Hard Prompts | Arena (LMArena) | 1518† | 1449† |
| LiveBench 2.0 Reasoning | LiveBench | 89.3%† | – |
| LiveBench 1.0 Reasoning | LiveBench | 89.3%† | – |
| GPQA Diamond (AA) | Artificial Analysis | 93.5% | 84.2% |
| Humanity's Last Exam (AA) | Artificial Analysis | 42.1% | 25.4% |
| HLE-Diamond (AA) | Artificial Analysis | 29.1% | – |
| GPQA Diamond (Vals) | Vals AI | 94.4%† | 85.6%† |
| Kagi LLM Benchmark | Kagi LLM Benchmark | – | 72.7%† |
| GPQA Diamond (HELM) | HELM Capabilities | – | 79.1%† |
| ARC-AGI-1 | ARC Prize | 97.5% | 56.2% |
| ARC-AGI-2 | ARC Prize | 82.9% | 7.5% |
| ARC-AGI-3 | ARC Prize | 24.2% | – |
| CritPt | Artificial Analysis | 12.3% | 0.0% |
| CritPt v2 hard subset (AA) | Artificial Analysis | 45.0% | 11.5%† |
| MysteryMechanism | Vals AI | 36.5%† | – |
| HLE-Diamond (Scale) | Scale AI SEAL | 34.3%† | – |
| Arena Expert | Arena (LMArena) | 1527† | 1445† |
| Chess Puzzles | Epoch AI Benchmarking Hub | 58.0%† | 32.0% |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 47.0%† | 15.0% |
| LMCA | Epoch AI Benchmarking Hub | 52.9% | 40.0%† |
| DTBench | Epoch AI Benchmarking Hub | 94.9% | 90.7%† |
| ForecastBench | ForecastBench | – | 65.0%† |
Show all 229 benchmarks
| Metric | Publisher | Gemini 3.8 Flash (medium) | GPT-5 (medium) |
|---|---|---|---|
| Coding | |||
| SWE-bench Verified (Epoch) | Epoch AI Benchmarking Hub | – | 71.5% |
| SciCode | SciCode | 55.1% | 42.9%† |
| WeirdML | WeirdML | 84.8%† | 39.8%† |
| FrontierCode | Cognition | 41.2% | – |
| GSO-Bench | GSO-Bench | – | 6.9%† |
| Arena Coding | Arena (LMArena) | 1531† | 1462† |
| Arena WebDev | Arena (LMArena) | 1584† | 1417 |
| LiveBench 2.0 Coding | LiveBench | 73.3%† | – |
| LiveBench 1.0 Coding | LiveBench | 72.5%† | – |
| SciCode (AA) | Artificial Analysis | 55.1% | – |
| LiveCodeBench | Vals AI | 89.5%† | 85.9%† |
| IOI | Vals AI | 56.9%† | – |
| SWE-bench (Vals) | Vals AI | 80.0%† | 69.0%† |
| SWE-Bench Pro | Scale AI SEAL | – | 41.8%† |
| Aider Polyglot | Aider polyglot leaderboard | – | 86.7% |
| SWE-bench Verified (bash only) | SWE-bench | – | 65.0% |
| SWE-bench Verified (any scaffold) | SWE-bench | – | 74.4%† |
| Code Migration | Vals AI | 36.5%† | – |
| Code Migration v1.1 | Vals AI | 38.6%† | – |
| ProgramBench | Vals AI | 1.0%† | – |
| SWE-bench Multimodal | Vals AI | 40.8%† | – |
| Vibe Code Bench 1-100 | Vals AI | 18.8%† | – |
| Vibe Code Bench v1.1 | Vals AI | 78.7%† | 20.1%† |
| SWE-Bench Pro (private) | Scale AI SEAL | – | 14.9%† |
| SWE Atlas v1.0: Codebase QnA | Scale AI SEAL | 47.0%† | – |
| SWE Atlas v1.0: Refactoring | Scale AI SEAL | 44.8%† | – |
| SWE Atlas v1.0: Test Writing | Scale AI SEAL | 53.7%† | – |
| SWE Atlas v1.1 | Scale AI SEAL | 57.3%† | – |
| SWE Atlas v1.1: Codebase QnA | Scale AI SEAL | 45.2%† | – |
| SWE Atlas v1.1: Test Writing | Scale AI SEAL | 71.1%† | – |
| SWE Atlas v1.1: Refactoring | Scale AI SEAL | 55.7%† | – |
| SWE-Bench Pro V2 | Scale AI SEAL | 94.9%† | – |
| SWE-Bench Pro V2 HARD | Scale AI SEAL | 58.8%† | – |
| MirrorCode | Epoch AI Benchmarking Hub | 38.9%† | 4.4%† |
| Surface Evolver Bench | Epoch AI Benchmarking Hub | 76.9%† | – |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | 71.0% | – |
| CursorBench | Cursor | 37.3% | – |
| ALE-Bench | Epoch AI Benchmarking Hub | 1269.7† | 1162.5† |
| AlgoTune | AlgoTune | – | 1.7† |
| FrontierSWE | Epoch AI Benchmarking Hub | 19.6%† | – |
| WeirdML v3 | WeirdML | 7.4%† | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 19.7% | – |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | 83.9% | 35.2%† |
| Startrise Frontend | Startrise | 64.6%† | – |
| GameBench 2 Overall | DuelLab GameBench 2 | 59.0† | – |
| GameBench 2 by reasoning setting | DuelLab GameBench 2 | 59.7 | 38.0 |
| Design Arena Website | Design Arena | 1291† | 1174† |
| FrontierCode 1.1 (AA) | Artificial Analysis | 37.1% | – |
| FrontierCode 2.0 (AA) | Artificial Analysis | 16.5% | – |
| LiveCodeBench (AA) | Artificial Analysis | – | 70.3% |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 18.2%† | 49.6% |
| GDPval | OpenAI | – | 34.8% |
| Remote Labor Index | Scale AI / CAIS | 5.8%† | 1.7%† |
| APEX-Agents | Mercor | 64.3%† | – |
| Arena Agent | Arena (LMArena) | 2.3† | – |
| LiveBench Long Agentic | LiveBench | 16.5%† | – |
| LiveBench Browser | LiveBench | 45.1%† | – |
| LiveBench Agentic Coding | LiveBench | 54.2%† | – |
| Terminal-Bench Hard | Artificial Analysis | – | 37.9% |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 86.5% |
| Terminal-Bench 2.1 (Vals) | Vals AI | 81.3%† | – |
| HiL-Bench | Scale AI SEAL | 41.5%† | – |
| GDPval-AA v2.1, normalised score | Artificial Analysis | 46.1% | – |
| GDPval-AA v2.1 | Artificial Analysis | 1432 | 925† |
| τ³-Banking (AA) | Artificial Analysis | 45.8% | 22.1%† |
| ITBench SRE (AA) | Artificial Analysis | 52.5%† | – |
| Analyst Agent (AA) | Artificial Analysis | 46.3%† | – |
| CUA-bench | Vals AI | 4.2%† | – |
| CyberBench | Vals AI | 43.8%† | – |
| Finance Agent v2 | Vals AI | 61.4%† | – |
| Finance Agent v2.1 | Vals AI | 65.8%† | – |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | 12.1%† | – |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 10.0%† | – |
| Legal Research Bench | Vals AI | 38.9%† | – |
| Poker Agent | Vals AI | – | 1103.2%† |
| SkillsBench | Vals AI | 58.0%† | – |
| SREBench | Vals AI | 11.1%† | – |
| Tax Agent Bench | Vals AI | 32.1%† | – |
| Terminal-Bench 2.0 (Vals) | Vals AI | – | 37.1%† |
| Terminal-Bench 4.0 (Vals) | Vals AI | 19.7%† | – |
| Terminal-Bench Science | Vals AI | 8.6%† | – |
| Time Horizon Index: KSP | Vals AI | 18.8%† | – |
| Arena Search | Arena (LMArena) | – | 1189† |
| Arena Agent Task Completion | Arena (LMArena) | 1.3† | – |
| BALROG | BALROG | – | 32.8%† |
| DeepResearch Bench | DeepResearch Bench | – | 48.6% |
| METR Time Horizons | METR | – | 69.6% |
| Vending-Bench 2 | Andon Labs | 5093.8† | – |
| AutomationBench | Artificial Analysis | 60.9% | – |
| GDP.pdf (AA) | Artificial Analysis | 22.8% | – |
| Harvey LAB-AA v1.1 | Harvey | 0.0%† | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1183 | – |
| Design Arena Agentic | Design Arena | 1225† | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 10.5% | – |
| CWE-Bench (AA) | Artificial Analysis | 52.5%† | – |
| DeepSec-Bench (AA) | Artificial Analysis | 14.5%† | – |
| CyberGym end-to-end (AA) | Artificial Analysis | 8.4%† | – |
| MKA (AA) | Artificial Analysis | 35.3%† | – |
| BrowseComp-Plus | BrowseComp-Plus | – | 57.6%† |
| BrowseComp | OpenAI | – | 20.1%† |
| DeepSearchQA | – | 59.4%† | |
| FACTS Search | 80.4%† | – | |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 68.4%† | 55.4%† |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 21.9%† | 21.9%† |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 98.9%† | 87.2% |
| MATH Level 5 | Epoch AI Benchmarking Hub | – | 97.9% |
| ProofBench | Vals AI | 48.0%† | 18.0%† |
| LiveBench Mathematics | LiveBench | 91.6%† | – |
| AIME (Vals) | Vals AI | – | 93.4%† |
| MGSM | Vals AI | – | 92.8%† |
| IMO 2025 | MathArena | – | 38.1%† |
| MathArena Apex | MathArena | – | 1.0%† |
| Omni-MATH (HELM) | HELM Capabilities | – | 64.7%† |
| MATH 500 | Vals AI | – | 96.0%† |
| Arena Maths | Arena (LMArena) | 1519† | 1411† |
| FrontierMath Erdős | Epoch AI Benchmarking Hub | 0.0%† | – |
| AIME 2025 (AA) | Artificial Analysis | – | 91.7% |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 69.7%† | 50.1%† |
| LiveBench 2.0 Data Analysis | LiveBench | 54.0%† | – |
| LiveBench Research | LiveBench | 79.4%† | – |
| LiveBench 1.0 Data Analysis | LiveBench | 54.0%† | – |
| AA-Omniscience v1.1 | Artificial Analysis | 23.7 | – |
| MMLU-Pro | Vals AI | 90.2%† | 86.5%† |
| LegalBench | Vals AI | 87.0%† | 86.0%† |
| CorpFin | Vals AI | – | 61.1%† |
| TaxEval | Vals AI | 74.5%† | 73.4%† |
| MedQA | Vals AI | – | 96.3%† |
| PRBench Finance | Scale AI SEAL | 49.5%† | 51.3%† |
| PRBench Legal | Scale AI SEAL | 48.9%† | 49.0%† |
| MultiNRC | Scale AI SEAL | – | 52.1%† |
| MMLU-Pro (HELM) | HELM Capabilities | – | 86.3%† |
| BioMysteryBench | Vals AI | 62.2%† | – |
| CaseLaw v2 | Vals AI | – | 66.5%† |
| Excel Modeling Benchmark | Vals AI | 72.2%† | – |
| Excel Modeling Benchmark v1.1 | Vals AI | 72.9%† | – |
| MedCode | Vals AI | 48.1%† | 49.6%† |
| MedScribe | Vals AI | 84.5%† | 83.7%† |
| Public Benefits Bench v1.1 | Vals AI | 65.3%† | – |
| Access to Justice Benchmark | Vals AI | 67.1%† | – |
| Financial Coaching Bench | Vals AI | 66.3%† | – |
| Immigration Bench | Vals AI | 63.4%† | – |
| EBR-bench | Epoch AI Benchmarking Hub | – | 12.7%† |
| GDP.pdf | Epoch AI Benchmarking Hub | 23.4% | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | 49.6% | – |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | 48.5% | – |
| Global-MMLU-Lite (AA) | Artificial Analysis | – | 90.4% |
| HealthBench Professional (AA) | OpenAI | 51.7%† | – |
| FACTS Parametric | 78.7%† | – | |
| Instruction following | |||
| LiveBench Language | LiveBench | 87.8%† | – |
| LiveBench Instruction Following | LiveBench | 81.4%† | – |
| IFBench | Artificial Analysis | – | 70.6% |
| MultiChallenge | Scale AI SEAL | – | 63.2%† |
| TutorBench | Scale AI SEAL | – | 55.3%† |
| IFEval (HELM) | HELM Capabilities | – | 87.5%† |
| WildBench (HELM) | HELM Capabilities | – | 85.7%† |
| Arena Instruction Following | Arena (LMArena) | 1492† | 1415† |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1497† | 1427† |
| EQ-Bench Creative Writing v3 | EQ-Bench | 1744† | 1626† |
| Arena Creative Writing | Arena (LMArena) | 1488† | 1389† |
| Arena Multi-turn | Arena (LMArena) | 1505† | 1449† |
| Arena Longer Queries | Arena (LMArena) | 1511† | 1435† |
| Lech Mazur Creative Writing | Lech Mazur | – | 86.0% |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1317† | 1232† |
| LiveBench Design | LiveBench | 69.7%† | – |
| MMMU-Pro | Artificial Analysis | 84.2% | 74.3% |
| VISTA | Scale AI SEAL | – | 49.7%† |
| MMMU-Pro (Vals) | Vals AI | 89.1%† | 81.5%† |
| MortgageTax | Vals AI | 65.3%† | 65.5%† |
| SAGE | Vals AI | 88.2%† | – |
| AudioMC | Scale AI SEAL | 60.4%† | – |
| AudioMC (text output) | Scale AI SEAL | 60.4%† | – |
| Humanity's Sixth Sense | Scale AI SEAL | 41.6%† | – |
| VTB | Scale AI SEAL | – | 17.0%† |
| GeoBench | GeoBench | 86.0%† | 81.0% |
| VPCT | Epoch AI Benchmarking Hub | – | 63.2% |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 38.6%† | – |
| Furniture Assembly | Epoch AI Benchmarking Hub | 31.7%† | – |
| VIALS-AA | Handshake AI | 41.0%† | – |
| Code2Video-AA | Artificial Analysis | 899† | – |
| AA-3D | Artificial Analysis | 974† | – |
| FACTS Multimodal | 57.6%† | – | |
| Long context | |||
| Fiction.LiveBench 120k | Fiction.live | – | 96.9% |
| AA-LCR v1.1 | Artificial Analysis | 84.0% | 76.0% |
| Arena Document | Arena (LMArena) | 1469† | – |
| MLCR | Artificial Analysis | 12.8% | – |
| FACTS Grounding | 68.2%† | – | |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 158.8† | 150† |
| LiveBench 2.0 | LiveBench | 57.3%† | – |
| LiveBench 1.0 | LiveBench | 75.8%† | – |
| AA Intelligence Index v5.0.1 | Artificial Analysis | 30.1 | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 39.8 | 22.9 |
| AA Progress Index v1 | Artificial Analysis | 319.1 | 264.4 |
| Vals Index | Vals AI | 56.0† | – |
| HELM Capabilities mean | HELM Capabilities | – | 80.7%† |
| Vals Finance Index | Vals AI | 60.5%† | – |
| Vals RSI Index | Vals AI | 19.9%† | – |
| Vals Science Index | Vals AI | 35.8%† | – |
| Vals Router Index | Vals AI | 56.0%† | – |
| Vals Legal Index | Vals AI | 69.8%† | – |
| AA Cyber Index | Artificial Analysis | 25.1† | – |
| Safety & honesty | |||
| FORTRESS | Scale AI SEAL | – | 17.0%† |
| MASK | Scale AI SEAL | – | 79.3%† |
| DistressBench | Scale AI SEAL | 70.7%† | – |
| Metadata | |||
| AA Openness Index | Artificial Analysis | – | 5.6 |
Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.
DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.
What changed
- Gemini 3.8 Flash: list price moves to $1.50 / $7.50 per 1M input / output tokens (announced).
- GPT-5 (medium): BenchLeader Index 58.2 → 59.0.
- GPT-5 (medium): 1 new result (Aider Polyglot).
- Gemini 3.8 Flash (medium): 1 score revision.
- GPT-5 (medium): 2 new results (Global-MMLU-Lite (AA), AIME 2025 (AA)); 1 score revision.
- GPT-5 (medium): 1 score revision.
- Gemini 3.8 Flash (medium): 2 score revisions.
- GPT-5 (medium): 1 score revision.
- Gemini 3.8 Flash (medium): 1 score revision.
- GPT-5 (medium): 1 score revision.
Related comparisons
- Gemini 3.8 Flash vs Gemini 4 Carbon
- GPT-5 vs Claude Opus 5.5
- Gemini 3 Pro vs GPT-5
- Gemini 3.8 Flash vs o3
- Gemini 3.8 Flash vs GPT-6.1 Sol
- Gemini 3.8 Flash vs Gemini 4 Argon
- Gemini 3.8 Flash vs Muse Wonder
- Gemini 3.8 Flash vs Muse Spark
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Gemini 3.8 Flash vs GPT-5: questions
- Is Gemini 3.8 Flash better than GPT-5?
- Gemini 3.8 Flash (medium) and GPT-5 (medium) are neck and neck on quality (60.2 vs 59.0, within the ±7.3 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
- Gemini 3.8 Flash vs GPT-5: which is cheaper, and will that change?
- Gemini 3.8 Flash is cheaper today: $1.50 against $3.44 per million tokens, blended at three input tokens per output token. Gemini 3.8 Flash's current price runs until 31 Dec 2026, then $3.00.
- Is Gemini 3.8 Flash or GPT-5 better for coding?
- GPT-5 scores higher in coding (57 vs 52 on the category index, where 50 is average).
- Which is faster, Gemini 3.8 Flash or GPT-5?
- Gemini 3.8 Flash streams faster (116 vs 82 output tokens per second, Artificial Analysis 7-day median).
- Is Gemini 3.8 Flash or GPT-5 better for agents and tool use?
- Gemini 3.8 Flash scores higher in agentic tasks (75 vs 58 on the category index, where 50 is average).
More questions (4)
- Which is better for long documents?
- GPT-5 scores higher on long-context benchmarks (70 vs 67). Gemini 3.8 Flash accepts more context: 1,048,576 against 400,000 tokens for GPT-5.
- Which has lower latency?
- Gemini 3.8 Flash answers first: 14.8 s against 22.9 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 3.8 Flash vs GPT-5?
- Across the 9 benchmarks both have, Gemini 3.8 Flash leads 9, GPT-5 leads 0. Every overall index that rates both puts Gemini 3.8 Flash ahead.
- When was Gemini 3.8 Flash released?
- Gemini 3.8 Flash was released on 2 Sep 2026.