GPT-6.1 Terravs.Gemini 4 Carbon
Which one is better?
Gemini 4 Carbon is clearly ahead in the overall index.
67
GPT-6.1 Terra
Rank 12 to 27
96
Gemini 4 Carbon
Rank 1 to 5
91
GPT-6.1 Terra
fairly slow
100
Gemini 4 Carbon
fairly fast
$2
GPT-6.1 Terra
fairly expensive
n/a
Gemini 4 Carbon
GPT-6.1 Terra
- Subscription
Subscriptions in Germany: n/a
- Via API
$1 input, $5 output per 1M tokens.
Gemini 4 Carbon
- Subscription
Subscriptions in Germany: n/a
- Via API
n/a (no API price known).
Overall index with data through Oct 29, 2026 · Prices from Oct 29, 2026 · Speed: measured by Artificial Analysis · Subscriptions as of Oct 24, 2026 · Subscription prices incl. VAT
Where is the difference?
Values from 0 to 100 per task area. We only name a winner when the gap is large enough and there are enough tests.
GPT-6.1 Terra ahead: 0Gemini 4 Carbon ahead: 6level: 0too little data: 1of 7 areas
AreaGPT-6.1 TerraGemini 4 Carbon0 to 100, higher is betterResult
Areas in the overall index
CodingGemini 4 Carbon aheadGPT-6.1 Terra: 84Gemini 4 Carbon: 92
ReasoningGemini 4 Carbon aheadGPT-6.1 Terra: 88Gemini 4 Carbon: 96
KnowledgeGemini 4 Carbon aheadGPT-6.1 Terra: 80Gemini 4 Carbon: above 99
AgentsGemini 4 Carbon aheadGPT-6.1 Terra: 57Gemini 4 Carbon: 88
Multimodalonly GPT-6.1 Terra has a scoreGPT-6.1 Terra: 70Gemini 4 Carbon: not enough data
Specialist fields
FinanceGemini 4 Carbon aheadGPT-6.1 Terra: 45Gemini 4 Carbon: 94
LegalGemini 4 Carbon aheadGPT-6.1 Terra: 8Gemini 4 Carbon: 94
See for yourself: same task, one attempt
The same prompt for both models. We show the latest attempt without fixes; it does not count toward the overall index.
Pelican on a bicycle
A community classic for free-form SVG drawing.
GPT-6.1 Terra
Not tested yet
Gemini 4 Carbon
Not tested yet
Prompt and test conditions
Generate an SVG of a pelican riding a bicycle
Token limit including reasoning: 8,192 tokens
Task origin (Simon Willison)All data side by side
The better value is bold. For the overall index only when the gap is clear.
| Metric | GPT-6.1 Terra | Gemini 4 Carbon |
|---|---|---|
| Intelligence | ||
| Overall index0 to 100, higher is better | 67 | 96 |
| Rank | Rank 12 to 27 of 95 | Rank 1 to 5 of 95 |
| Price per 1M tokens | ||
| Inputwhat you send | $1above 272,000 tokens: $2 | n/a |
| Outputwhat the model writes | $5above 272,000 tokens: $7.50 | n/a |
| Speed | ||
| Tokens per second | 91 | 100 |
| Technical | ||
| Context windowhow much text the model reads at once | 1.05 M tokens≈ 1,580 A4 pages | 1 M tokens≈ 1,500 A4 pages |
| Knowledge cutoff | Jun 1, 2026 | n/a |
| Released | Oct 26, 2026 | n/a |
| Availability | Active | Limited access |
| Open weights | no | no |
- GPT-6.1 Terra, 12-hour cache write: The table's cache-write rate covers 30-minute retention. With 12-hour retention it is $2.50 per 1M tokens through 272,000 input tokens, $5 above. Source
- GPT-6.1 Terra, Long-context billing: Above 272,000 input tokens, the higher rates apply to the entire request. Input and cache rates double, while output costs 1.5 times the standard rate. Source
- GPT-6.1 Terra, Batch and Flex pricing: 50% off standard pricing Source
- GPT-6.1 Terra, Fast pricing: Fast mode costs twice the standard price Source
- GPT-6.1 Terra, Regional processing: 10% surcharge over standard processing Source
- Gemini 4 Carbon, Announced introductory pricing: $2 input and $10 output per 1M tokens. Cached input receives a 95% discount, giving $0.10 per 1M tokens. This announcement does not establish an already available public API. Source
- Gemini 4 Carbon, Announced pricing after the introductory period: Afterward, $4 input and $20 output per 1M tokens. Google has not specified when the introductory period ends. Source
- Gemini 4 Carbon, Artificial Analysis pricing assumptions: Artificial Analysis calculates test costs using $4 input, $20 output, and $0.20 cached input per 1M tokens. These are the evaluator’s pricing assumptions, not confirmation of a publicly available API tariff. Source
Prices without caching or batch discounts · Page count as a rule of thumb from Artificial Analysis
Shared benchmarks
Single results from the same benchmark version, task and metric. A value is bold only with the same test setup and a clear gap.
A: GPT-6.1 TerraB: Gemini 4 Carbon
AA-BriefcaseOverall · higher is betterGPT-6.1 Terra: 1,263.5 eloGemini 4 Carbon: 1,609.49 eloArtificial AnalysisAs of 10/29/2026
AA-OmniscienceOverall · higher is betterGPT-6.1 Terra: 27.767 pointsGemini 4 Carbon: 53.5 pointsArtificial AnalysisAs of 10/29/2026
AA-Omniscience AccuracyOverall · higher is betterGPT-6.1 Terra: 47.3%Gemini 4 Carbon: 58.3%Artificial AnalysisAs of 10/29/2026
AA-Omniscience Hallucination RateOverall · lower is betterGPT-6.1 Terra: 37.0%Gemini 4 Carbon: 11.4%Artificial AnalysisAs of 10/29/2026
Artificial Analysis Intelligence IndexOverall · higher is betterGPT-6.1 Terra: 42.334 pointsGemini 4 Carbon: 59.426 pointsArtificial AnalysisAs of 10/29/2026
AutomationBench-AAOverall · higher is betterGPT-6.1 Terra: 58.5%Gemini 4 Carbon: 77.8%Artificial AnalysisAs of 10/29/2026
CritPt v2Overall · higher is betterGPT-6.1 Terra: 25.7%Gemini 4 Carbon: 36.0%Artificial AnalysisAs of 10/29/2026
CritPt v2 (Hard Subset)Overall · higher is betterGPT-6.1 Terra: 60.5%Gemini 4 Carbon: 63.0%Artificial AnalysisAs of 10/29/2026
FrontierCodeOverall · higher is betterGPT-6.1 Terra: 37.5%Gemini 4 Carbon: 40.6%Artificial AnalysisAs of 10/29/2026
FrontierCode 1.1 MainOverall · higher is betterGPT-6.1 Terra: 37.2%Gemini 4 Carbon: 48.5%Artificial AnalysisAs of 10/29/2026
Show 14 more benchmarks
GDP.pdfOverall · higher is betterGPT-6.1 Terra: 28.0%Gemini 4 Carbon: 28.4%Artificial AnalysisAs of 10/29/2026
GDPval-AAOverall · higher is betterGPT-6.1 Terra: 1,353.15 eloGemini 4 Carbon: 1,733.07 eloArtificial AnalysisAs of 10/29/2026
HLE-DiamondOverall · higher is betterGPT-6.1 Terra: 35.4%Gemini 4 Carbon: 69.1%Artificial AnalysisAs of 10/29/2026
Terminal-BenchOverall · higher is betterGPT-6.1 Terra: 41.9%Gemini 4 Carbon: 65.2%Artificial AnalysisAs of 10/29/2026
Terminal-Bench-ScienceOverall · higher is betterGPT-6.1 Terra: 33.3%Gemini 4 Carbon: 64.8%Artificial AnalysisAs of 10/29/2026
Vals IndexOverall · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 60.5%Gemini 4 Carbon: 70.4%Vals AIAs of 10/27/2026
Vals IndexCode Migration v1.1 · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 72.2%Gemini 4 Carbon: 80.1%Vals AIAs of 10/27/2026
Vals IndexEMB v1.1 · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 73.3%Gemini 4 Carbon: 76.5%Vals AIAs of 10/27/2026
Vals IndexFinance Agent v2.1 · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 56%Gemini 4 Carbon: 67.2%Vals AIAs of 10/27/2026
Vals IndexHLAB-Vals-1.1 · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 1.7%Gemini 4 Carbon: 17.5%Vals AIAs of 10/27/2026
Vals IndexLegal Research · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 26%Gemini 4 Carbon: 55.8%Vals AIAs of 10/27/2026
Vals IndexTax Agent · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 15.2%Gemini 4 Carbon: 48.9%Vals AIAs of 10/27/2026
Vals IndexTerminal-Bench 4.0 · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 49.5%Gemini 4 Carbon: 63.1%Vals AIAs of 10/27/2026
Vals IndexVibe Code Bench v1.1 · higher is betterTest setup not documented as identicalGPT-6.1 Terra: 84%Gemini 4 Carbon: 89.1%Vals AIAs of 10/27/2026
Compare another matchup
More matchups
Related LLM comparisons
Compare GPT-6.1 Terra and Gemini 4 Carbon with other leading models.