Head-to-head · Updated Nov 3, 2026
Claude Sonnet 5.5 vs GPT-6.1 Sol
Anthropic’s Claude Sonnet 5.5 and OpenAI’s GPT-6.1 Sol on independent tests for professional work, cost and speed.
Short answer
Claude Sonnet 5.5 leads on 8 of 13 quality scores, while GPT-6.1 Sol has the higher Intelligence Index.
- FinanceGPT-6.1 Sol56.2 vs 55.8
- Legal workGPT-6.1 Sol53.7 vs 52.2
- HealthcareClaude Sonnet 5.554.2 vs 52.0
- EngineeringGPT-6.1 Sol59.3 vs 58.2
- Strategy & operationsClaude Sonnet 5.558.1 vs 57.6
- Coding agentsClaude Sonnet 5.563.6% vs 56.1%
- Expert knowledgeClaude Sonnet 5.555.0% vs 52.9%
- Fewer made-up answersClaude Sonnet 5.543.3% vs 51.9%
- Lower costGPT-6.1 Sol$1,124 vs $7,981
- SpeedClaude Sonnet 5.5140 tok/s vs 93 tok/s
Scores, cost and speed
| Measure | Claude Sonnet 5.5 (Max, Default Fallback) | GPT-6.1 Sol (Max) |
|---|---|---|
| Overall | ||
| Intelligence IndexCombined score across nine evaluations | 54.2 | Leads: 54.5 |
| Professional work | ||
| Finance & accounting | 55.8 | Leads: 56.2 |
| Legal | 52.2 | Leads: 53.7 |
| Healthcare & medical | Leads: 54.2 | 52.0 |
| Engineering | 58.2 | Leads: 59.3 |
| Strategy & operations | Leads: 58.1 | 57.6 |
| Economics | 56.0 | Leads: 57.4 |
| Specific tests | ||
| Terminal-Bench 4.0Agentic coding tasks | Leads: 63.6% | 56.1% |
| FrontierCode 2.0Hard programming problems | Leads: 41.7% | 40.6% |
| Humanity’s Last ExamExpert-level questions | Leads: 55.0% | 52.9% |
| HealthBenchHealth conversations graded by physicians’ rubrics | Leads: 62.1% | 58.5% |
| GDPval-AAKnowledge-work deliverables, Elo | Leads: 1838 | 1609 |
| Hallucination rateAA-Omniscience; lower is better | Leads: 43.3% | 51.9% |
| Cost & speed | ||
| Input priceUS$ per 1M tokens | $2 | $2 |
| Output priceUS$ per 1M tokens | $10 | $10 |
| Cost to run the indexUS$, includes reasoning tokens | $7,981 | Leads: $1,124 |
| Output speedMedian tokens per second | Leads: 140 tok/s | 93 tok/s |
| Context windowTokens | 1M | 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.
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.