Skip to content
Model Pareto

Compare models

Side by side: quality per category, price, speed, and every benchmark cell with where it came from. Pick up to four.
GPT-6 SolQwen3.8 Max (0902)Step 5 Preview
Focus

Coding: quality vs price

Compared models are ringed; the other 100 ranked here are greyed.

Lower price is better. Pareto frontier: Granite 4.2 3B, NVIDIA Nemotron 3 Nano 30B A3B, gpt-oss-20b, Ling-3.0-flash-VL, MiMo-V2.6-Flash, GLM-5.3 Flash, DeepSeek V4.1 Flash, MiMo-V2.6-Pro, Claude Sonnet 5.5, Claude Opus 5.5. 26 models are not plotted: Xing4.0-29B-A4B, Nex-N2.5-Pro, Atria Dawn Preview, Intern-S2-397B, Solar Open2 250B, Phi-4-reasoning-plus, Motif 3, Nex-N2-Pro, Nex-N2.5-Mini, A.X-K2, K2 Horizon 375B A23B, K-EXAONE 2.0 0803, K2 Horizon MoVA 36B A4B, Intern-S2-Preview (35B-A3B), Command A+, North Mini Code, G9v3-39A5B, Falcon-H1-34B-Instruct, Devstral 2, K2 Horizon 7B, Devstral Small 2, MiniCPM5-2B, Ling 3.0 Tiny, K2 Horizon 3.7B, K2 Horizon 0.9B, LFM2.5-2.6B.

26 models with no price data — shown in the strip at the left edge

Xing4.0-29B-A4B, Nex-N2.5-Pro, Atria Dawn Preview, Intern-S2-397B, Solar Open2 250B, Phi-4-reasoning-plus, Motif 3, Nex-N2-Pro, Nex-N2.5-Mini, A.X-K2, K2 Horizon 375B A23B, K-EXAONE 2.0 0803, K2 Horizon MoVA 36B A4B, Intern-S2-Preview (35B-A3B), Command A+, North Mini Code, G9v3-39A5B, Falcon-H1-34B-Instruct, Devstral 2, K2 Horizon 7B, Devstral Small 2, MiniCPM5-2B, Ling 3.0 Tiny, K2 Horizon 3.7B, K2 Horizon 0.9B, LFM2.5-2.6B

  • Best-value frontier (nothing is both cheaper and better)
  • Evidencestrong → weak
  • Estimated from other categories
  • GPT-6 Sol
    Quality
    78.7
    Rank
    #6/103
    Price
    $4.00/M tok
    Speed
    110tok/s
    In$2.00Out$10/M tok

    vs Step 5 Preview: +5.0 quality · 2.8× price

  • Qwen3.8 Max (0902)
    Quality
    59.1
    Rank
    #24/103
    Price
    $3.00/M tok
    Speed
    39tok/s
    In$2.00Out$6.00/M tok

    vs GPT-6 Sol: −19.6 quality · 0.75× price

  • Step 5 Preview
    Quality
    73.7
    Rank
    #10/103
    Price
    $1.43/M tok
    Speed
    85tok/s
    In$1.00▾Out$2.70▾/M tok

    vs GPT-6 Sol: −5.0 quality · 0.36× price

Overview

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
Lab
OpenAI
Alibaba Qwen
StepFun
Released
Sep 22, 2026
Sep 2, 2026
Sep 20, 2026
Weights
Proprietary
Proprietary
Proprietary
Input price
$/M tok
$2.00
$2.00
$1.00Best
Output price
$/M tok
$10
$6.00
$2.70Best
Blended price
$/M tok · 3:1 in:out
$4.00
$3.00
$1.425Best
Output speed
tok/s
110Best
39
85
Context
tokens
872K
1MBest
1MBest

Quality by category

0–100 within each category (100 = best tracked model). Rank is among all models ranked there. Click a row to focus it.

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
Coding
78.7±4.5Best
#6/103Verified
59.1±4.5
#24/103Verified
73.7±6.4
#10/103Mixed
Writing & chat
83.1±4.3Best
#7/146Verified
82.4±6.0
#9/146Verified
82.6±12.7
#8/146Estimated from other categoriesbest value
Vision
86.8±4.8Best
#4/82Verifiedbest value
77.4±4.8
#10/82Verified
66.1±6.3
#20/82Verified
Hard reasoning
85.8±5.1Best
#8/182Verified
82.1±3.7
#13/182Verified
83.2±4.0
#10/182Mixed
Agents
76.0±7.8
#17/173Verified
85.2±5.8Best
#7/173Verified
75.4±6.3
#18/173Mixed

Coding benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
SWE-bench Verified
day 0
~#2/14~89.3%Estimated
  1. #1Claude Opus 596.0%
  2. ~#2GPT-6 SolEstimated89.3%
  3. #2Claude Sonnet 585.2%
estimated from related benchmarks
~#2/14~86.2%Estimated
  1. #1Claude Opus 596.0%
  2. ~#2Qwen3.8 Max (0902)Estimated86.2%
  3. #2Claude Sonnet 585.2%
estimated from related benchmarks
~#2/14~88.2%Estimated
  1. #1Claude Opus 596.0%
  2. ~#2Step 5 PreviewEstimated88.2%
  3. #2Claude Sonnet 585.2%
estimated from related benchmarks
SWE-bench Pro (Public)
day 0
~#5/19~71.9%Estimated
  1. #1Claude Opus 5.589.9%
  2. #4Claude Opus 579.2%
  3. ~#5GPT-6 SolEstimated71.9%
  4. #5Intern-S2-397B68.5%
estimated from related benchmarks
~#6/19~66.8%Estimated
  1. #1Claude Opus 5.589.9%
  2. #5Intern-S2-397B68.5%
  3. ~#6Qwen3.8 Max (0902)Estimated66.8%
  4. #6Hy4 preview65.7%
estimated from related benchmarks
~#5/19~70.2%Estimated
  1. #1Claude Opus 5.589.9%
  2. #4Claude Opus 579.2%
  3. ~#5Step 5 PreviewEstimated70.2%
  4. #5Intern-S2-397B68.5%
estimated from related benchmarks
DeepSWE v1.1
day 0
#7/3069.0%BestVerified
  1. #1DeepSeek V4.1 Flash74.2%
  2. #6GPT-5.6 Terra69.6%
  3. #7GPT-6 Sol69.0%
  4. #8Claude Opus 5.568.4%
artificialanalysis.ai · 2026-09-24
#27/3051.0%Verified
  1. #1DeepSeek V4.1 Flash74.2%
  2. #26Nex-N2.5-Pro55.8%
  3. #27Qwen3.8 Max (0902)51.0%
  4. #28Qwen3.8 27B42.2%
artificialanalysis.ai · 2026-09-24
#11/3067.7%Lab-reported
  1. #1DeepSeek V4.1 Flash74.2%
  2. #10MiMo-V2.6-Flash67.9%
  3. #11Step 5 Preview67.7%
  4. #12GPT-6 Astra67.6%
stepfun.com · 2026-09-20
Terminal-Bench 4.0
day 0
#6/8343.9%BestVerified
  1. #1Claude Sonnet 5.563.6%
  2. #5Claude Opus 549.0%
  3. #6GPT-6 Sol43.9%
  4. #7GLM-5.341.9%
artificialanalysis.ai · 2026-09-24
#8/8338.9%Verified
  1. #1Claude Sonnet 5.563.6%
  2. #7GLM-5.341.9%
  3. #8Qwen3.8 Max (0902)38.9%
  4. #9GPT-5.6 Terra35.4%
artificialanalysis.ai · 2026-09-24
#11/8333.3%Verified
  1. #1Claude Sonnet 5.563.6%
  2. #11Muse Spark 1.333.3%
  3. #11Step 5 Preview33.3%
  4. #13GLM-5.3 Flash32.8%
artificialanalysis.ai · 2026-09-24
LiveCodeBench
third-party
~#2/20~92.1%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. ~#2GPT-6 SolEstimated92.1%
  3. #2Qwen3.8 Flash-Next91.9%
estimated from related benchmarks
~#2/20~92.0%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. ~#2Qwen3.8 Max (0902)Estimated92.0%
  3. #2Qwen3.8 Flash-Next91.9%
estimated from related benchmarks
~#2/20~92.1%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. ~#2Step 5 PreviewEstimated92.1%
  3. #2Qwen3.8 Flash-Next91.9%
estimated from related benchmarks
SciCode
third-party
#10/9557.6%Verified
  1. #1Claude Opus 5.566.9%
  2. #9Gemini 3.1 Pro58.7%
  3. #10GPT-6 Sol57.6%
  4. #11Grok 4.757.4%
artificialanalysis.ai · 2026-09-24
#22/9552.1%Verified
  1. #1Claude Opus 5.566.9%
  2. #22GPT-5.4 mini52.1%
  3. #22Qwen3.8 Max (0902)52.1%
  4. #24DeepSeek V4.1 Flash51.9%
artificialanalysis.ai · 2026-09-24
#7/9558.9%BestVerified
  1. #1Claude Opus 5.566.9%
  2. #6GLM-5.359.0%
  3. #7Step 5 Preview58.9%
  4. #8Muse Spark 1.358.8%
artificialanalysis.ai · 2026-09-24
AA Coding Agent Index
third-party
#5/1456.7BestVerified
  1. #1Claude Opus 5.566.0
  2. #4Claude Opus 559.7
  3. #5GPT-6 Sol56.7
  4. #6Grok 4.756.3
artificialanalysis.ai · 2026-09-24
#11/1443.3Verified
  1. #1Claude Opus 5.566.0
  2. #10Grok 4.647.0
  3. #11Qwen3.8 Max (0902)43.3
  4. #12DeepSeek V4 Pro (0813)43.1
artificialanalysis.ai · 2026-09-24
~#9/15~53.2Estimated
  1. #1Claude Opus 5.566.0
  2. #8GLM-5.353.6
  3. ~#9Step 5 PreviewEstimated53.2
  4. #9Kimi K351.9
estimated from related benchmarks
Benchmarks in other categories (4)

Writing & chat benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
IFBench
day 0
~#7/126~80.7%Estimated
  1. #1MiniMax-M382.9%
  2. #6Intern-S2-Preview (35B-A3B)80.9%
  3. ~#7GPT-6 SolEstimated80.7%
  4. #7Nemotron Cascade 2 30B A3B80.4%
estimated from related benchmarks
~#7/126~80.8%Estimated
  1. #1MiniMax-M382.9%
  2. #6Intern-S2-Preview (35B-A3B)80.9%
  3. ~#7Qwen3.8 Max (0902)Estimated80.8%
  4. #7Nemotron Cascade 2 30B A3B80.4%
estimated from related benchmarks
~#3/126~82.7%Estimated
  1. #1MiniMax-M382.9%
  2. #2Qwen3.8 2.4T A95B82.8%
  3. ~#3Step 5 PreviewEstimated82.7%
  4. #3Qwen3.8-Omni-Flash81.5%
estimated from related benchmarks
MMMLU
day 0
~#2/5~90.4%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2GPT-6 SolEstimated90.4%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
~#2/5~91.3%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2Qwen3.8 Max (0902)Estimated91.3%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
~#2/5~91.2%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2Step 5 PreviewEstimated91.2%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
LMArena Text, Non-English (Elo)
third-party
#16/531448Verified
  1. #1Claude Opus 5.51501
  2. #16DeepSeek V4 Pro (0813)1448
  3. #16GPT-6 Sol1448
  4. #18Claude Sonnet 51447
lmarena.ai · 2026-09-25
#9/531467BestVerified
  1. #1Claude Opus 5.51501
  2. #8GPT-5.51474
  3. #9Qwen3.8 Max (0902)1467
  4. #10GLM-5.31466
lmarena.ai · 2026-09-25
~#9/54~1469Estimated
  1. #1Claude Opus 5.51501
  2. #8GPT-5.51474
  3. ~#9Step 5 PreviewEstimated1469
  4. #9Qwen3.8 Max (0902)1467
estimated from related benchmarks
LMArena Text (Elo)
third-party
#18/531457Verified
  1. #1Claude Opus 5.51509
  2. #17Claude Sonnet 51461
  3. #18GPT-6 Sol1457
  4. #19Gemini 3.5 Flash-Lite1456
lmarena.ai · 2026-09-25
#9/531481BestVerified
  1. #1Claude Opus 5.51509
  2. #8GPT-5.51482
  3. #9Qwen3.8 Max (0902)1481
  4. #10GLM-5.31480
lmarena.ai · 2026-09-24
~#13/54~1479Estimated
  1. #1Claude Opus 5.51509
  2. #10MiMo-V2.6-Pro1480
  3. ~#13Step 5 PreviewEstimated1479
  4. #13DeepSeek V4.1 Flash1477
estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)
third-party
#4/252125Verified
  1. #1GPT-6 Astra2173
  2. #3Claude Opus 52133
  3. #4GPT-6 Sol2125
  4. #5Kimi K32082
eqbench.com · 2026-09-24
~#9/26~1958Estimated
  1. #1GPT-6 Astra2173
  2. #8Grok 4.72007
  3. ~#9Qwen3.8 Max (0902)Estimated1958
  4. #9Muse Spark 1.31906
estimated from related benchmarks
~#9/26~1972Estimated
  1. #1GPT-6 Astra2173
  2. #8Grok 4.72007
  3. ~#9Step 5 PreviewEstimated1972
  4. #9Muse Spark 1.31906
estimated from related benchmarks

Vision benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
MMMU
day 0
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2GPT-6 SolEstimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2Qwen3.8 Max (0902)Estimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2Step 5 PreviewEstimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
MMMU-Pro
day 0
#5/7683.3%BestVerified
  1. #1Claude Opus 5.587.7%
  2. #4Claude Opus 584.7%
  3. #5GPT-6 Sol83.3%
  4. #6Qwen3.8 Max (0902)82.8%
artificialanalysis.ai · 2026-09-24
#6/7682.8%Verified
  1. #1Claude Opus 5.587.7%
  2. #5GPT-6 Sol83.3%
  3. #6Qwen3.8 Max (0902)82.8%
  4. #7Gemini 3.1 Pro82.4%
artificialanalysis.ai · 2026-09-24
#25/7676.4%Verified
  1. #1Claude Opus 5.587.7%
  2. #24Grok Build 0.1 061676.5%
  3. #25Step 5 Preview76.4%
  4. #26Qwen3.8 27B76.3%
artificialanalysis.ai · 2026-09-24
CharXiv Reasoning
day 0
~#2/9~85.0%Estimated
  1. #1Gemini 3.8 Flash86.2%
  2. ~#2GPT-6 SolEstimated85.0%
  3. #2Kimi K384.8%
estimated from related benchmarks
~#4/9~84.1%Estimated
  1. #1Gemini 3.8 Flash86.2%
  2. #3Qwen3.8 Flash-Next84.6%
  3. ~#4Qwen3.8 Max (0902)Estimated84.1%
  4. #4Qwen3.8-Omni-Flash83.5%
estimated from related benchmarks
~#5/9~81.3%Estimated
  1. #1Gemini 3.8 Flash86.2%
  2. #4Qwen3.8-Omni-Flash83.5%
  3. ~#5Step 5 PreviewEstimated81.3%
  4. #5Muse Glimmer78.8%
estimated from related benchmarks
OmniDocBench v1.5
third-party
~#4/8~90.8%Estimated
  1. #1MiniMax-M391.6%
  2. #2Qwen3.8 27B91.1%
  3. ~#4GPT-6 SolEstimated90.8%
  4. #4Gemini 3.1 Pro85.3%
estimated from related benchmarks
~#4/8~90.7%Estimated
  1. #1MiniMax-M391.6%
  2. #2Qwen3.8 27B91.1%
  3. ~#4Qwen3.8 Max (0902)Estimated90.7%
  4. #4Gemini 3.1 Pro85.3%
estimated from related benchmarks
~#4/8~88.1%Estimated
  1. #1MiniMax-M391.6%
  2. #2Qwen3.8 27B91.1%
  3. ~#4Step 5 PreviewEstimated88.1%
  4. #4Gemini 3.1 Pro85.3%
estimated from related benchmarks
GDP.pdf (AA)
third-party
#5/4224.8%BestVerified
  1. #1GPT-6 Astra31.0%
  2. #3Claude Opus 5.526.2%
  3. #5GPT-6 Sol24.8%
  4. #6GPT-5.6 Terra24.0%
artificialanalysis.ai · 2026-09-24
#7/4222.8%Verified
  1. #1GPT-6 Astra31.0%
  2. #6GPT-5.6 Terra24.0%
  3. #7Qwen3.8 Max (0902)22.8%
  4. #8Kimi K322.0%
artificialanalysis.ai · 2026-09-24
#20/4214.8%Verified
  1. #1GPT-6 Astra31.0%
  2. #19GLM-5.3 Flash15.4%
  3. #20Step 5 Preview14.8%
  4. #21Gemini 3.5 Flash-Lite13.6%
artificialanalysis.ai · 2026-09-24
Chartography
third-party
#3/2253.6%BestVerified
  1. #1GPT-6 Astra71.0%
  2. #2Claude Opus 5.566.3%
  3. #3GPT-6 Sol53.6%
  4. #4Claude Fable 5.146.2%
surgehq.ai · 2026-09-24
#8/2229.1%Verified
  1. #1GPT-6 Astra71.0%
  2. #8GPT-6 Luna29.1%
  3. #8Qwen3.8 Max (0902)29.1%
  4. #10Muse Spark 1.327.6%
surgehq.ai · 2026-09-24
~#14/23~25.1%Estimated
  1. #1GPT-6 Astra71.0%
  2. #13Gemini 3.1 Pro26.1%
  3. ~#14Step 5 PreviewEstimated25.1%
  4. #14Grok 4.620.5%
estimated from related benchmarks

Hard reasoning benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
GPQA Diamond
day 0
~#3/173~94.7%Estimated
  1. #1GPT-6 Astra96.1%
  2. #2Gemini 3.8 Flash95.3%
  3. ~#3GPT-6 SolEstimated94.7%
  4. #3Gemini 3.1 Pro94.1%
estimated from related benchmarks
#13/17292.8%Verified
  1. #1GPT-6 Astra96.1%
  2. #13DeepSeek V4 Pro (0813)92.8%
  3. #13Qwen3.8 Max (0902)92.8%
  4. #15GPT-5.6 Terra92.5%
artificialanalysis.ai · 2026-09-24
#5/17293.5%BestLab-reported
  1. #1GPT-6 Astra96.1%
  2. #5Qwen3.8 2.4T A95B93.5%
  3. #5Step 5 Preview93.5%
  4. #11Claude Opus 593.2%
stepfun.com · 2026-09-20
Humanity's Last Exam
day 0
#8/17847.9%BestVerified
  1. #1Claude Opus 5.561.4%
  2. #7Muse Spark 1.348.7%
  3. #8GPT-6 Sol47.9%
  4. #9Gemini 3.8 Flash47.8%
artificialanalysis.ai · 2026-09-24
#16/17843.1%Verified
  1. #1Claude Opus 5.561.4%
  2. #16Grok 4.743.1%
  3. #16Qwen3.8 Max (0902)43.1%
  4. #18GPT-5.6 Terra42.9%
artificialanalysis.ai · 2026-09-24
#12/17846.5%Verified
  1. #1Claude Opus 5.561.4%
  2. #11Kimi K346.9%
  3. #12Step 5 Preview46.5%
  4. #13GPT-5.545.8%
artificialanalysis.ai · 2026-09-24
AIME (latest)
day 0
~#2/25~96.5%Estimated
  1. #1Inkling97.1%
  2. ~#2GPT-6 SolEstimated96.5%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
~#2/25~96.8%Estimated
  1. #1Inkling97.1%
  2. ~#2Qwen3.8 Max (0902)Estimated96.8%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
~#2/25~96.8%Estimated
  1. #1Inkling97.1%
  2. ~#2Step 5 PreviewEstimated96.8%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
AA Intelligence Index
third-party
#7/7447.5BestVerified
  1. #1Claude Opus 5.557.6
  2. #6Muse Spark 1.348.1
  3. #7GPT-6 Sol47.5
  4. #8Grok 4.746.4
artificialanalysis.ai · 2026-09-24
#10/7445.4Verified
  1. #1Claude Opus 5.557.6
  2. #9MiMo-V2.6-Pro46.3
  3. #10Qwen3.8 Max (0902)45.4
  4. #11GLM-5.344.8
artificialanalysis.ai · 2026-09-24
#13/7443.7Verified
  1. #1Claude Opus 5.557.6
  2. #12Grok 4.644.2
  3. #13Step 5 Preview43.7
  4. #14Kimi K343.6
artificialanalysis.ai · 2026-09-24

Agents benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
GPT-6 Sol
Qwen3.8 Max (0902)
Step 5 Preview
τ²-bench (Telecom)
day 0
~#2/116~98.2%Estimated
  1. #1Step 3.7 Flash98.5%
  2. ~#2GPT-6 SolEstimated98.2%
  3. #2Gemini 3.1 Pro95.6%
estimated from related benchmarks
~#2/116~98.3%Estimated
  1. #1Step 3.7 Flash98.5%
  2. ~#2Qwen3.8 Max (0902)Estimated98.3%
  3. #2Gemini 3.1 Pro95.6%
estimated from related benchmarks
~#2/116~98.1%Estimated
  1. #1Step 3.7 Flash98.5%
  2. ~#2Step 5 PreviewEstimated98.1%
  3. #2Gemini 3.1 Pro95.6%
estimated from related benchmarks
OSWorld-Verified
day 0
~#3/12~82.8%Estimated
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. ~#3GPT-6 SolEstimated82.8%
  4. #3Nex-N2.5-Pro82.2%
estimated from related benchmarks
~#3/12~83.5%Estimated
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. ~#3Qwen3.8 Max (0902)Estimated83.5%
  4. #3Nex-N2.5-Pro82.2%
estimated from related benchmarks
~#3/12~82.3%Estimated
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. ~#3Step 5 PreviewEstimated82.3%
  4. #3Nex-N2.5-Pro82.2%
estimated from related benchmarks
OSWorld 2.0
day 0
~#3/5~54.2%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3GPT-6 SolEstimated54.2%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
~#3/5~57.5%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Qwen3.8 Max (0902)Estimated57.5%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
~#3/5~53.8%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Step 5 PreviewEstimated53.8%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
BrowseComp
day 0
~#7/18~88.6%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #6Step 5 Preview88.7%
  3. ~#7GPT-6 SolEstimated88.6%
  4. #7GPT-5.6 Terra87.5%
estimated from related benchmarks
~#5/18~90.2%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #4Claude Opus 590.8%
  3. ~#5Qwen3.8 Max (0902)Estimated90.2%
  4. #5Nex-N2.5-Pro89.7%
estimated from related benchmarks
#6/1788.7%Lab-reported
  1. #1Atria Dawn Preview92.5%
  2. #5Nex-N2.5-Pro89.7%
  3. #6Step 5 Preview88.7%
  4. #7GPT-5.6 Terra87.5%
stepfun.com · 2026-09-20
GDPval-AA (Elo)
third-party
#17/861487Verified
  1. #1Claude Opus 5.51846
  2. #16Kimi K31524
  3. #17GPT-6 Sol1487
  4. #18Claude Sonnet 51449
artificialanalysis.ai · 2026-09-24
#7/861668BestVerified
  1. #1Claude Opus 5.51846
  2. #6MiMo-V2.6-Pro1673
  3. #7Qwen3.8 Max (0902)1668
  4. #8GLM-5.31646
artificialanalysis.ai · 2026-09-24
#14/861566Verified
  1. #1Claude Opus 5.51846
  2. #13Qwen3.8 2.4T A95B1597
  3. #14Step 5 Preview1566
  4. #15GPT-6 Astra1542
artificialanalysis.ai · 2026-09-24
AutomationBench-AA
third-party
#7/7961.6%BestVerified
  1. #1Claude Opus 5.569.5%
  2. #6GLM-5.362.2%
  3. #7GPT-6 Sol61.6%
  4. #8GLM-5.3 Flash60.4%
artificialanalysis.ai · 2026-09-24
#18/7956.2%Verified
  1. #1Claude Opus 5.569.5%
  2. #17Claude Opus 556.6%
  3. #18Qwen3.8 Max (0902)56.2%
  4. #19Qwen3.8 Flash-Next55.9%
artificialanalysis.ai · 2026-09-24
#21/7951.0%Verified
  1. #1Claude Opus 5.569.5%
  2. #20GPT-6 Luna53.2%
  3. #21Step 5 Preview51.0%
  4. #22Qwen3.8 27B48.2%
artificialanalysis.ai · 2026-09-24
τ-Bench Banking (AA)
third-party
~#16/89~39.5%Estimated
  1. #1Muse Spark 1.350.5%
  2. #15DeepSeek V4 Pro (0813)39.6%
  3. ~#16GPT-6 SolEstimated39.5%
  4. #16GPT-5.539.0%
estimated from related benchmarks
#5/8847.8%Verified
  1. #1Muse Spark 1.350.5%
  2. #4Qwen3.8 27B48.0%
  3. #5Qwen3.8 Max (0902)47.8%
  4. #6Claude Fable 5.147.2%
artificialanalysis.ai · 2026-09-24
~#17/89~38.6%Estimated
  1. #1Muse Spark 1.350.5%
  2. #16GPT-5.539.0%
  3. ~#17Step 5 PreviewEstimated38.6%
  4. #17Claude Sonnet 537.3%
estimated from related benchmarks
Not comparable across labs (1)
OSWorld 2.0 (offline set v2026.08.08)
day 0
#2/364.4%Lab-reported
  1. #1GPT-6 Astra72.6%
  2. #2GPT-6 Sol64.4%
  3. #3GPT-6 Luna52.7%
openai.com · 2026-09-22
—
—