OpenAIproprietary
Released: 2026-09-29Artificial Analysis

GPT-6.1 Sol

OpenRouter ranked #10 foundation model processing 3.89T tokens weekly. Deployed by openai.

Intelligence Index
54.6/ 100
Artificial Analysis Intelligence Index (independent composite)
Throughput Speed
49tok/sec
Output streaming throughput
Time to First Answer
191s
Time to first answer token, including any thinking
Cost to Run the Index
1.27USD
Artificial Analysis cost of the full Index run

Measured results · Artificial Analysis

Independent evaluations of GPT-6.1 Sol (Max configuration). Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.

Intelligence Index
54.6
Cost to run the Index
$1.27
Time per task
22.9 min
Omniscience index
42.0
Reasoning model
Yes
Cached-input price
$0.1 / 1M

Agentic work

Multimodal & science

Reasoning-effort variants evaluated

VariantIntelligence IndexCost to run IndexOutput speed (tok/s)
Max54.6$1.2749.4
Xhigh52.5$0.7650.2
High50.3$0.5948.2
Medium46.3$0.3650.6
Low34.8$0.1649.7
Full analysis on Artificial Analysis

Headline Evaluation Breakdown

Six independently measured evaluations (Artificial Analysis)
Reasoning · CritPt v2 (Hard)69.5%
Coding & agents · Terminal-Bench 4.056.1%
Long context · AA-LCR83.0%
Knowledge · Humanity's Last Exam52.9%

Sourced Benchmark Evaluations (22)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
AA-Omniscience (Non-Hallucination)Active48.1%
93
2026-11-15independentSource ↗
VIALS-AAActive59%
93
2026-11-15independentSource ↗
Harvey LAB-AAActive6.9%
93
2026-11-15independentSource ↗
AA-AnalystAgentActive50%
93
2026-11-15independentSource ↗
AutomationBench-AAActive64.9%
93
2026-11-15independentSource ↗
Multimodal Knowledge Agent (MKA)Active47.7%
93
2026-11-15independentSource ↗
MLCR-AA (Medical Long Context Reasoning)Active38.3%
93
2026-11-15independentSource ↗
GDP.pdfActive31%
93
2026-11-15independentSource ↗
FrontierCode-AA 2.0Active40.6%
93
2026-11-15independentSource ↗
HealthBench ProfessionalActive65.1%
93
2026-11-15independentSource ↗
HealthBenchActive58.5%
93
2026-11-15independentSource ↗
FrontierCodeActive49.1%
98
2026-11-15independentSource ↗
AA-Omniscience (Accuracy)Active61.8%
93
2026-11-15independentSource ↗
HLE-DiamondActive58.8%
93
2026-11-15independentSource ↗
CritPt v2 (Hard)Active69.5%
93
2026-11-15independentSource ↗
CritPtActive31.7%
93
2026-11-15independentSource ↗
AA-LCR (Long Context Reasoning)Active83%
93
2026-11-15independentSource ↗
Terminal-Bench-Science 0.1Active58.1%
93
2026-11-15independentSource ↗
Terminal-Bench 4.0Active56.1%
93
2026-11-15independentSource ↗
SciCodeActive54.2%
93
2026-11-15independentSource ↗
MMMU-ProActive86%
93
2026-11-15independentSource ↗
Humanity's Last ExamActive52.9%
93
2026-11-15independentSource ↗
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI

client = OpenAI()
response = client.chat.completions.create(
    model="gpt-6-1-sol",
    messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
    temperature=0.2
)
print(response.choices[0].message.content)

Evaluation & Sourcing Notes

Scores listed for GPT-6.1 Sol represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.