Gemini 4 Hydrogen vs GPT-5
Gemini 4 Hydrogen leads GPT-5 on coding, 77.9 to 38.8 on the Vibe Coding Index, and Gemini 4 Hydrogen costs 1.3x less per output token ($7.50 vs $10.00 per million). Updated November 14, 2026; 32 benchmarks have a result for both.
vs boards won of 32 shared
Gemini 4 Hydrogen or GPT-5: which to pick
Pick Gemini 4 Hydrogen for coding: it scores 77.9 on the Vibe Coding Index against 38.8 for GPT-5.
GPT-5 is neither cheaper nor larger in context, and leads no benchmark both models were measured on.
Model AGemini 4 HydrogenGoogle77.9/1004 independent benchmarksNo cost per solved taskModel BGPT-5OpenAI38.8/1003 independent benchmarks$0.39 per solved task
- Shared results
- 32
- Gemini 4 Hydrogen only
- 18
- GPT-5 only
- 86
- Like-for-like categories
- 6 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
Gemini 4 Hydrogen
Gemini 4 Hydrogen scores 77.9 against 38.8.
4 and 3 benchmarks behind the scores
Reasoning
GPQA Diamond
No call
Only GPT-5 has a GPQA Diamond result, 85.6%.
Directional only
Math
AIME
No call
Only GPT-5 has a AIME result, 93.4%.
Directional only
Knowledge
MMLU Pro
No call
Only GPT-5 has a MMLU Pro result, 86.5%.
Directional only
Web design
Design Arena: websites
Gemini 4 Hydrogen
Gemini 4 Hydrogen leads, 1320 against 1176.
Same board, both measured
General
Vals Index
No call
Only Gemini 4 Hydrogen has a Vals Index result, 68.5%.
Directional only
Long documents
Listed context window
Gemini 4 Hydrogen
Gemini 4 Hydrogen takes 1.05M tokens against 400K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemini 4 Hydrogen
Gemini 4 Hydrogen $0.0053, GPT-5 $0.0063 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only GPT-5 has the number, $0.39.
Directional only
At a glance
- Coding index
- Gemini 4 Hydrogen 77.9 from 4 benchmarks, GPT-5 38.8 from 3 benchmarks.
- Widest category
- Coding, 33.3 points apart on average over 5 shared benchmarks.
- Widest single gap
- Vibe Code Bench (Vals AI): Gemini 4 Hydrogen 87.8% against GPT-5 20.1%, 67.7 points apart on a 0 to 100 scale.
- Output price
- Gemini 4 Hydrogen $7.50 against GPT-5 $10.00 per million output tokens, 1.3x apart.
- Context window
- Gemini 4 Hydrogen takes 1.05M tokens, GPT-5 400K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Gemini 4 Hydrogen | Bars | GPT-5 | Ties |
|---|---|---|---|---|
| Coding | 5 | 0 | ||
| Reasoning | 4 | 0 | ||
| Math | 2 | 0 | ||
| Knowledge | 4 | 0 | ||
| Web design | 4 | 0 | ||
| General | 13 | 0 | ||
| All shared boards | 32 | 0 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index50
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| Aider polyglot | Not measured | 88.0% | Only GPT-5 measured |
| SWE-bench Verified | Not measured | 74.4% | Only GPT-5 measured |
| LiveBench coding | 78.5% | Not measured | Only Gemini 4 Hydrogen measured |
| LMArena WebDev | 1704 | 1417 | Gemini 4 Hydrogen |
| AA Coding Index | Not measured | 37.8 | Only GPT-5 measured |
| LiveCodeBench (Vals AI) | Not measured | 85.9% | Only GPT-5 measured |
| SWE-bench Verified (Vals AI) | Not measured | 69.0% | Only GPT-5 measured |
| Terminal-Bench 4 (Vals AI) | 54.0% | Not measured | Only Gemini 4 Hydrogen measured |
| Vibe Code Bench (Vals AI) | 87.8% | 20.1% | Gemini 4 Hydrogen |
| LiveCodeBench (Artificial Analysis run) | Not measured | 84.6% | Only GPT-5 measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 37.9% | Only GPT-5 measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 54.0% | Not measured | Only Gemini 4 Hydrogen measured |
| FrontierCode 2.0 | 39.0% | Not measured | Only Gemini 4 Hydrogen measured |
| Terminal-Bench | Not measured | 49.6% | Only GPT-5 measured |
| GSO-Bench | Not measured | 6.9% | Only GPT-5 measured |
| WeirdML | Not measured | 60.7% | Only GPT-5 measured |
| MirrorCode | Not measured | 4.4% | Only GPT-5 measured |
| ALE-Bench | Not measured | 807.6 | Only GPT-5 measured |
| LMArena Text: coding | 1516 | 1436 | Gemini 4 Hydrogen |
| LMArena Text: software and IT | 1513 | 1420 | Gemini 4 Hydrogen |
| LMArena WebDev: React | 1724 | Not measured | Only Gemini 4 Hydrogen measured |
| LMArena WebDev: HTML | 1743 | 1425 | Gemini 4 Hydrogen |
| SciCode | Not measured | 0.0% | Only GPT-5 measured |
| GameBench 2 | Not measured | 46.2 | Only GPT-5 measured |
| Terminal-Bench 2.0 | Not measured | 49.6% | Only GPT-5 measured |
| CursorBench 4.0 | 57.8% | Not measured | Only Gemini 4 Hydrogen measured |
| Neither model has a result on the other 26 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| Epoch Capabilities Index | Not measured | 150.0 | Only GPT-5 measured |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index14 boards, none shared
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| LiveBench research | 86.8% | Not measured | Only Gemini 4 Hydrogen measured |
| LiveBench browser | 77.1% | Not measured | Only Gemini 4 Hydrogen measured |
| LiveBench long agentic | 61.8% | Not measured | Only Gemini 4 Hydrogen measured |
| Tax Agent Bench (Vals AI) | 52.3% | Not measured | Only Gemini 4 Hydrogen measured |
| Legal Research Bench (Vals AI) | 53.8% | Not measured | Only Gemini 4 Hydrogen measured |
| Harvey's Legal Agent Benchmark (Vals AI) | 11.3% | Not measured | Only Gemini 4 Hydrogen measured |
| τ²-Bench Telecom | Not measured | 86.5% | Only GPT-5 measured |
| METR time horizons | Not measured | 69.6% | Only GPT-5 measured |
| Remote Labor Index | Not measured | 1.7% | Only GPT-5 measured |
| GDPval | Not measured | 34.8% | Only GPT-5 measured |
| BALROG | Not measured | 32.8% | Only GPT-5 measured |
| DeepResearch Bench | Not measured | 49.6% | Only GPT-5 measured |
| LMArena Search | Not measured | 1133 | Only GPT-5 measured |
| BrowseComp-Plus | Not measured | 80.0% | Only GPT-5 measured |
| Neither model has a result on the other 30 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 85.6% | Only GPT-5 measured |
| LiveBench reasoning | 88.6% | Not measured | Only Gemini 4 Hydrogen measured |
| MMMU Pro (Vals AI) | Not measured | 81.5% | Only GPT-5 measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 85.4% | Only GPT-5 measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 28.5% | Only GPT-5 measured |
| HLE-Diamond (Artificial Analysis run) | 58.8% | Not measured | Only Gemini 4 Hydrogen measured |
| CritPt v2, hard subset (Artificial Analysis run) | 66.0% | 11.5% | Gemini 4 Hydrogen |
| GPQA Diamond (Epoch AI run) | Not measured | 86.2% | Only GPT-5 measured |
| ARC-AGI | Not measured | 65.7% | Only GPT-5 measured |
| ARC-AGI-2 | Not measured | 9.9% | Only GPT-5 measured |
| SimpleBench | Not measured | 56.7% | Only GPT-5 measured |
| Chess puzzles | Not measured | 36.0% | Only GPT-5 measured |
| Mystery game puzzles | Not measured | 23.0% | Only GPT-5 measured |
| EBR-bench | Not measured | 12.7% | Only GPT-5 measured |
| VPCT | Not measured | 66.0% | Only GPT-5 measured |
| LMArena Text: hard prompts | 1520 | 1416 | Gemini 4 Hydrogen |
| LMArena Text: expert | 1558 | 1419 | Gemini 4 Hydrogen |
| LMArena Text: hard prompts in English | 1514 | 1419 | Gemini 4 Hydrogen |
| LMArena Vision: diagrams | Not measured | 1251 | Only GPT-5 measured |
| LMArena Vision: homework | Not measured | 1279 | Only GPT-5 measured |
| Kagi LLM Benchmark | Not measured | 72.7% | Only GPT-5 measured |
| Neither model has a result on the other 5 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| AIME | Not measured | 93.4% | Only GPT-5 measured |
| MATH 500 (Vals AI) | Not measured | 96.0% | Only GPT-5 measured |
| MGSM (Vals AI) | Not measured | 92.8% | Only GPT-5 measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 94.3% | Only GPT-5 measured |
| OTIS Mock AIME | Not measured | 91.4% | Only GPT-5 measured |
| FrontierMath, tiers 1 to 3 | Not measured | 55.4% | Only GPT-5 measured |
| FrontierMath, tier 4 | Not measured | 22.0% | Only GPT-5 measured |
| ProofBench | Not measured | 18.0% | Only GPT-5 measured |
| MATH level 5 | Not measured | 98.1% | Only GPT-5 measured |
| LMArena Text: math | 1507 | 1408 | Gemini 4 Hydrogen |
| LMArena Text: mathematical work | 1534 | 1405 | Gemini 4 Hydrogen |
| MathArena | Not measured | 16.9% | Only GPT-5 measured |
| Neither model has a result on the other 2 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 86.5% | Only GPT-5 measured |
| LegalBench (Vals AI) | Not measured | 86.0% | Only GPT-5 measured |
| CaseLaw (Vals AI) | Not measured | 66.5% | Only GPT-5 measured |
| CorpFin (Vals AI) | Not measured | 61.1% | Only GPT-5 measured |
| TaxEval (Vals AI) | Not measured | 73.4% | Only GPT-5 measured |
| MortgageTax (Vals AI) | Not measured | 65.5% | Only GPT-5 measured |
| MedQA (Vals AI) | Not measured | 96.3% | Only GPT-5 measured |
| MedCode (Vals AI) | Not measured | 49.6% | Only GPT-5 measured |
| MMLU Pro (Artificial Analysis run) | Not measured | 87.1% | Only GPT-5 measured |
| AA-Omniscience | 46.3 | Not measured | Only Gemini 4 Hydrogen measured |
| SimpleQA Verified | Not measured | 50.1% | Only GPT-5 measured |
| GeoBench | Not measured | 81.0% | Only GPT-5 measured |
| LMArena Text: science | 1519 | 1416 | Gemini 4 Hydrogen |
| LMArena Text: medicine and healthcare | 1488 | 1413 | Gemini 4 Hydrogen |
| LMArena Text: legal and government | 1532 | 1430 | Gemini 4 Hydrogen |
| LMArena Text: business and finance | 1486 | 1399 | Gemini 4 Hydrogen |
| PRBench Finance | Not measured | 51.3% | Only GPT-5 measured |
| PRBench Legal | Not measured | 49.0% | Only GPT-5 measured |
| Neither model has a result on the other 3 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| Design Arena: websites | 1320 | 1176 | Gemini 4 Hydrogen |
| Design Arena: UI components | Not measured | 1170 | Only GPT-5 measured |
| Design Arena: game dev | 1369 | 1174 | Gemini 4 Hydrogen |
| Design Arena: data viz | Not measured | 1218 | Only GPT-5 measured |
| Design Arena: 3D | 1415 | 1035 | Gemini 4 Hydrogen |
| Design Arena: all code categories | 1333 | 1157 | Gemini 4 Hydrogen |
| Design Arena: SVG | Not measured | 1178 | Only GPT-5 measured |
| Design Arena: ASCII art | Not measured | 1136 | Only GPT-5 measured |
| LiveBench design | 77.3% | Not measured | Only Gemini 4 Hydrogen measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index130
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| Vals Index | 68.5% | Not measured | Only Gemini 4 Hydrogen measured |
| LiveBench data analysis | 83.2% | Not measured | Only Gemini 4 Hydrogen measured |
| MedScribe (Vals AI) | Not measured | 83.7% | Only GPT-5 measured |
| IFBench | Not measured | 73.1% | Only GPT-5 measured |
| Long context reasoning (AA-LCR) | Not measured | 78.2% | Only GPT-5 measured |
| GDPval-AA | Not measured | 925 | Only GPT-5 measured |
| DTBench | Not measured | 90.7% | Only GPT-5 measured |
| LMCA | Not measured | 40.0% | Only GPT-5 measured |
| Fiction.LiveBench | Not measured | 97.2% | Only GPT-5 measured |
| LMArena Text | 1502 | 1406 | Gemini 4 Hydrogen |
| LMArena Text: instruction following | 1519 | 1388 | Gemini 4 Hydrogen |
| LMArena Text: creative writing | 1509 | 1365 | Gemini 4 Hydrogen |
| LMArena Vision | Not measured | 1232 | Only GPT-5 measured |
| LMArena Text: English | 1499 | 1409 | Gemini 4 Hydrogen |
| LMArena Text: non-English | 1495 | 1397 | Gemini 4 Hydrogen |
| LMArena Text: Chinese | 1631 | 1422 | Gemini 4 Hydrogen |
| LMArena Text: German | Not measured | 1416 | Only GPT-5 measured |
| LMArena Text: Japanese | Not measured | 1409 | Only GPT-5 measured |
| LMArena Text: Korean | Not measured | 1360 | Only GPT-5 measured |
| LMArena Text: French | 1512 | 1408 | Gemini 4 Hydrogen |
| LMArena Text: Spanish | 1494 | 1400 | Gemini 4 Hydrogen |
| LMArena Text: Russian | 1492 | 1407 | Gemini 4 Hydrogen |
| LMArena Text: Polish | Not measured | 1431 | Only GPT-5 measured |
| LMArena Text: multi-turn | 1485 | 1426 | Gemini 4 Hydrogen |
| LMArena Text: longer queries | 1524 | 1398 | Gemini 4 Hydrogen |
| LMArena Text: writing and language | 1500 | 1384 | Gemini 4 Hydrogen |
| LMArena Text: entertainment and media | 1503 | 1379 | Gemini 4 Hydrogen |
| LMArena Vision: English | Not measured | 1231 | Only GPT-5 measured |
| LMArena Vision: Chinese | Not measured | 1267 | Only GPT-5 measured |
| LMArena Vision: reading text | Not measured | 1245 | Only GPT-5 measured |
| LMArena Vision: humor | Not measured | 1218 | Only GPT-5 measured |
| LMArena Vision: captioning | Not measured | 1213 | Only GPT-5 measured |
| LMArena Vision: recognizing things | Not measured | 1261 | Only GPT-5 measured |
| LMArena Vision: creative writing | Not measured | 1211 | Only GPT-5 measured |
| MultiChallenge | Not measured | 63.2% | Only GPT-5 measured |
| Neither model has a result on the other 14 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemini 4 Hydrogen | GPT-5 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Gemini 4 Hydrogen$0.0053Fits in one request
GPT-5$0.0063Fits in one request
Gemini 4 Hydrogen costs less
Repository review
50K input + 3K output tokens
Gemini 4 Hydrogen$0.098Fits in one request
GPT-5$0.092Fits in one request
GPT-5 costs less
Agent loop
200K input + 10K output tokens
Gemini 4 Hydrogen$0.375Fits in one request
GPT-5$0.350Fits in one request
GPT-5 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Google Gemini 4 Hydrogen | OpenAI GPT-5 |
|---|---|---|
| Maker | OpenAI | |
| Weights | No public weights listed | No public weights listed |
| Released | November 12, 2026 | August 7, 2025 |
| Context window | 1.05M | 400K |
| Input $/Mtok | $1.50 | $1.25 |
| Output $/Mtok | $7.50 | $10.00 |
| Coding index | 77.9 | 38.8 |
| Independent measurements behind it | 4 | 3 |
| Cost per solved task | Not computable | $0.39 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemini 4 Hydrogen better than GPT-5 for coding?
Pick Gemini 4 Hydrogen for coding: it scores 77.9 on the Vibe Coding Index against 38.8 for GPT-5. GPT-5 is neither cheaper nor larger in context, and leads no benchmark both models were measured on.
Which is cheaper, Gemini 4 Hydrogen or GPT-5?
Gemini 4 Hydrogen costs $1.50 per million input tokens and $7.50 per million output tokens; GPT-5 costs $1.25 and $10.00. Cost per solved task is known for GPT-5 only.
Which has the bigger context window, Gemini 4 Hydrogen or GPT-5?
Gemini 4 Hydrogen, with 1.05M tokens against 400K.
How many benchmarks is the Gemini 4 Hydrogen vs GPT-5 comparison based on?
32 benchmarks have a result for both models as of November 14, 2026. Gemini 4 Hydrogen's coding index rests on 4 independent benchmarks and GPT-5's on 3. Where only one model has a result, the page says so instead of naming a winner.