Head-to-head · Updated Nov 3, 2026
Grok vs Gemini
The highest-scoring Grok and Gemini models, Grok 4.7 and Gemini 4 Carbon, on independent tests for professional work, cost and speed.
Short answer
Gemini 4 Carbon leads on 10 of 10 quality scores, including the overall Intelligence Index.
- FinanceGemini 4 Carbon62.6 vs 51.7
- Legal workGemini 4 Carbon62.0 vs 47.2
- EngineeringGemini 4 Carbon64.7 vs 41.6
- Strategy & operationsGemini 4 Carbon60.9 vs 54.8
- Coding agentsGemini 4 Carbon65.2% vs 25.8%
- Fewer made-up answersGemini 4 Carbon11.4% vs 20.0%
- Lower costGrok 4.7$5,905 vs $8,240
- SpeedGemini 4 Carbon94 tok/s vs 86 tok/s
Scores, cost and speed
| Measure | Grok 4.7 (Xhigh) | Gemini 4 Carbon (High)Limited early access |
|---|---|---|
| Overall | ||
| Intelligence IndexCombined score across nine evaluations | 35.3 | Leads: 59.4 |
| Professional work | ||
| Finance & accounting | 51.7 | Leads: 62.6 |
| Legal | 47.2 | Leads: 62.0 |
| Healthcare & medical | 46.6 | — |
| Engineering | 41.6 | Leads: 64.7 |
| Strategy & operations | 54.8 | Leads: 60.9 |
| Economics | 47.6 | Leads: 61.0 |
| Specific tests | ||
| Terminal-Bench 4.0Agentic coding tasks | 25.8% | Leads: 65.2% |
| FrontierCode 2.0Hard programming problems | 28.7% | Leads: 40.6% |
| Humanity’s Last ExamExpert-level questions | 43.1% | — |
| HealthBenchHealth conversations graded by physicians’ rubrics | 56.0% | — |
| GDPval-AAKnowledge-work deliverables, Elo | 1718 | Leads: 1733 |
| Hallucination rateAA-Omniscience; lower is better | 20.0% | Leads: 11.4% |
| Cost & speed | ||
| Input priceUS$ per 1M tokens | Leads: $2 | $4 |
| Output priceUS$ per 1M tokens | Leads: $6 | $20 |
| Cost to run the indexUS$, includes reasoning tokens | Leads: $5,905 | $8,240 |
| Output speedMedian tokens per second | 86 tok/s | Leads: 94 tok/s |
| Context windowTokens | 500K | Leads: 1M |
Scores, prices and speeds: Artificial Analysis, independently measured, checked Nov 3, 2026. Model names include the tested reasoning setting. Stubbi has not rerun these evaluations. Consumer apps such as ChatGPT, Claude and Gemini may use other models and add features like search and memory.
All measured Grok and Gemini models
| Model and setting | Released | Index | Price in / out |
|---|---|---|---|
| Gemini 4 Carbon (High)Google · Limited early access | Oct 2026 | 59.4 | $4 / $20 |
| Grok 4.7 (Xhigh)SpaceXAI | Sep 2026 | 35.3 | $2 / $6 |
| Gemini 3.8 Flash (High)Google | Sep 2026 | 30.5 | $0.75 / $3.75 |
Test them on your own work
Public scores can’t show how each model handles your documents, clients or standards. Tell us the task, and we’ll consider a side-by-side comparison with the outputs you can judge.