Skip to content
Model Pareto

GPT-6 Sol

OpenAIProprietaryopenai.com ↗
Compare with nearest rivalsCompare with…
Input
$2.00/M tok
Output
$10/M tok
Blended
$4.00/M tok · 3:1
Speed
110tok/s
Context
872Ktokens

Rankings

Benchmark scores

~ italic, dashed = no published score yet, estimated from related benchmarks.

Coding

SWE-bench VerifiedEstimated
~#2 of 14 (estimated)~89.3%
  1. #1Claude Opus 596.0%
  2. ~#2GPT-6 SolEstimated89.3%
  3. #2Claude Sonnet 585.2%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#5 of 19 (estimated)~71.9%
  1. #1Claude Opus 5.589.9%
  2. #2Claude Sonnet 5.581.3%
  3. #4Claude Opus 579.2%
  4. ~#5GPT-6 SolEstimated71.9%
  5. #5Intern-S2-397B68.5%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#2 of 20 (estimated)~92.1%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. ~#2GPT-6 SolEstimated92.1%
  3. #2Qwen3.8 Flash-Next91.9%
No published score — estimated from related benchmarks

Writing & chat

IFBenchEstimated
~#7 of 126 (estimated)~80.7%
  1. #1MiniMax-M382.9%
  2. #2Qwen3.8 2.4T A95B82.8%
  3. #6Intern-S2-Preview (35B-A3B)80.9%
  4. ~#7GPT-6 SolEstimated80.7%
  5. #7Nemotron Cascade 2 30B A3B80.4%
No published score — estimated from related benchmarks
MMMLUEstimated
~#2 of 5 (estimated)~90.4%
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2GPT-6 SolEstimated90.4%
  3. #2Gemma 4 31B88.4%
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Verified
#16 of 531448 Elo
  1. #1Claude Opus 5.51501
  2. #2Claude Fable 5.1Few votes1496
  3. #16DeepSeek V4 Pro (0813)1448
  4. #16GPT-6 Sol1448
  5. #18Claude Sonnet 51447
lmarena.ai · 2026-09-25
LMArena Text (Elo)Verified
#18 of 531457 Elo
  1. #1Claude Opus 5.51509
  2. #2Claude Fable 5.11501
  3. #17Claude Sonnet 51461
  4. #18GPT-6 Sol1457
  5. #19Gemini 3.5 Flash-Lite1456
lmarena.ai · 2026-09-25

Vision

MMMUEstimated
~#2 of 3 (estimated)~71.3%
  1. #1Claude Haiku 4.573.2%
  2. ~#2GPT-6 SolEstimated71.3%
  3. #2Llama 4 Maverick69.4%
No published score — estimated from related benchmarks
CharXiv ReasoningEstimated
~#2 of 9 (estimated)~85.0%
  1. #1Gemini 3.8 Flash86.2%
  2. ~#2GPT-6 SolEstimated85.0%
  3. #2Kimi K384.8%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~#4 of 8 (estimated)~90.8%
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #2Qwen3.8 27B91.1%
  4. ~#4GPT-6 SolEstimated90.8%
  5. #4Gemini 3.1 Pro85.3%
No published score — estimated from related benchmarks

Hard reasoning

GPQA DiamondEstimated
~#3 of 173 (estimated)~94.7%
  1. #1GPT-6 Astra96.1%
  2. #2Gemini 3.8 Flash95.3%
  3. ~#3GPT-6 SolEstimated94.7%
  4. #3Gemini 3.1 Pro94.1%
No published score — estimated from related benchmarks
AIME (latest)Estimated
~#2 of 25 (estimated)~96.5%
  1. #1Inkling97.1%
  2. ~#2GPT-6 SolEstimated96.5%
  3. #2Muse Glimmer94.7%
No published score — estimated from related benchmarks

Agents

τ²-bench (Telecom)Estimated
~#2 of 116 (estimated)~98.2%
  1. #1Step 3.7 Flash98.5%
  2. ~#2GPT-6 SolEstimated98.2%
  3. #2Gemini 3.1 Pro95.6%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#3 of 12 (estimated)~82.8%
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. ~#3GPT-6 SolEstimated82.8%
  4. #3Nex-N2.5-Pro82.2%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~#3 of 5 (estimated)~54.2%
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3GPT-6 SolEstimated54.2%
  4. #3GPT-5.6 Terra50.2%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#7 of 18 (estimated)~88.6%
  1. #1Atria Dawn Preview92.5%
  2. #2GPT-6 Astra91.5%
  3. #6Step 5 Preview88.7%
  4. ~#7GPT-6 SolEstimated88.6%
  5. #7GPT-5.6 Terra87.5%
No published score — estimated from related benchmarks
τ-Bench Banking (AA)Estimated
~#16 of 89 (estimated)~39.5%
  1. #1Muse Spark 1.350.5%
  2. #2GLM-5.350.3%
  3. #15DeepSeek V4 Pro (0813)39.6%
  4. ~#16GPT-6 SolEstimated39.5%
  5. #16GPT-5.539.0%
No published score — estimated from related benchmarks
Not comparable across labs (1)

Kept for reference, not counted in rankings: these use a lab-specific task set, answer key or scoring, so scores can't be compared fairly across labs. Only published scores are shown.