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.
DeepSeek V4.1 FlashKimi K3Qwen3.8 Flash-Next
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
  • DeepSeek V4.1 Flash
    Quality
    66.4
    Rank
    #14/103
    Price
    $0.52/M tok
    Speed
    232tok/s
    In$0.30Out$1.20/M tok

    vs Kimi K3: +1.2 quality · 0.09× price

  • Kimi K3
    Quality
    65.2
    Rank
    #16/103
    Price
    $6.00/M tok
    Speed
    37tok/s
    In$3.00Out$15/M tok

    vs DeepSeek V4.1 Flash: −1.2 quality · 11× price

  • Qwen3.8 Flash-Next
    Quality
    62.0
    Rank
    #20/103
    Price
    $0.23/M tok
    Speed
    54tok/s
    In$0.15▾Out$0.47▾/M tok

    vs DeepSeek V4.1 Flash: −4.5 quality · 0.44× price

Overview

Metric
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
Lab
DeepSeek
Moonshot AI (Kimi)
Alibaba Qwen
Released
Sep 10, 2026
Jul 16, 2026
Aug 26, 2026
Weights
Open
Open
Open
Input price
$/M tok
$0.30
$3.00
$0.15Best
Output price
$/M tok
$1.20
$15
$0.47Best
Blended price
$/M tok · 3:1 in:out
$0.525
$6.00
$0.23Best
Output speed
tok/s
232Best
37
54
Context
tokens
1M
1MBest
262K

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
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
Coding
66.4±6.4Best
#14/103Mixedbest value
65.2±4.5
#16/103Verified
62.0±5.2
#20/103Mixed
Writing & chat
69.8±4.3
#26/146Verified
89.1±4.3Best
#5/146Verifiedbest value
80.0±9.7
#12/146Lab-reportedbest value
Vision
56.4±4.8
#35/82Verified
77.0±3.9Best
#11/82Mixed
72.6±5.3
#16/82Mixed
Hard reasoning
76.1±4.0
#23/182Mixed
83.4±3.7Best
#9/182Verified
76.0±3.7
#24/182Verifiedbest value
Agents
82.1±7.8
#13/173Verified
82.4±3.5Best
#12/173Mixed
81.7±5.8
#14/173Verifiedbest value

Coding benchmarks

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

Metric
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
SWE-bench Verified
day 0
~#3/14~83.2%Estimated
  1. #1Claude Opus 596.0%
  2. #2Claude Sonnet 585.2%
  3. ~#3DeepSeek V4.1 FlashEstimated83.2%
  4. #3Gemini 3.1 Pro80.6%
estimated from related benchmarks
~#2/14~85.3%Estimated
  1. #1Claude Opus 596.0%
  2. ~#2Kimi K3Estimated85.3%
  3. #2Claude Sonnet 585.2%
estimated from related benchmarks
~#3/14~82.9%Estimated
  1. #1Claude Opus 596.0%
  2. #2Claude Sonnet 585.2%
  3. ~#3Qwen3.8 Flash-NextEstimated82.9%
  4. #3Gemini 3.1 Pro80.6%
estimated from related benchmarks
SWE-bench Pro (Public)
day 0
~#6/19~65.9%Estimated
  1. #1Claude Opus 5.589.9%
  2. #5Intern-S2-397B68.5%
  3. ~#6DeepSeek V4.1 FlashEstimated65.9%
  4. #6Hy4 preview65.7%
estimated from related benchmarks
~#6/19~66.6%Estimated
  1. #1Claude Opus 5.589.9%
  2. #5Intern-S2-397B68.5%
  3. ~#6Kimi K3Estimated66.6%
  4. #6Hy4 preview65.7%
estimated from related benchmarks
~#7/19~64.7%Estimated
  1. #1Claude Opus 5.589.9%
  2. #6Hy4 preview65.7%
  3. ~#7Qwen3.8 Flash-NextEstimated64.7%
  4. #7GPT-5.6 Terra63.4%
estimated from related benchmarks
DeepSWE v1.1
day 0
#1/3074.2%BestLab-reported
  1. #1DeepSeek V4.1 Flash74.2%
  2. #2Grok 4.772.6%
huggingface.co · 2026-09-10
#8/3068.4%Verified
  1. #1DeepSeek V4.1 Flash74.2%
  2. #8Claude Opus 5.568.4%
  3. #8Kimi K368.4%
  4. #10MiMo-V2.6-Flash67.9%
artificialanalysis.ai · 2026-09-24
#22/3058.7%Lab-reported
  1. #1DeepSeek V4.1 Flash74.2%
  2. #21GLM-5.361.4%
  3. #22Qwen3.8 Flash-Next58.7%
  4. #23Qwen3.8-Omni-Flash57.8%
huggingface.co · 2026-08-26
Terminal-Bench 4.0
day 0
#14/8326.8%BestVerified
  1. #1Claude Sonnet 5.563.6%
  2. #13GLM-5.3 Flash32.8%
  3. #14DeepSeek V4.1 Flash26.8%
  4. #15Grok 4.725.8%
artificialanalysis.ai · 2026-09-24
#23/8312.6%Verified
  1. #1Claude Sonnet 5.563.6%
  2. #23GPT-6 Luna12.6%
  3. #23Kimi K312.6%
  4. #26Qwen3.8 2.4T A95B11.1%
artificialanalysis.ai · 2026-09-24
#16/8325.3%Verified
  1. #1Claude Sonnet 5.563.6%
  2. #15Grok 4.725.8%
  3. #16Qwen3.8 Flash-Next25.3%
  4. #17MiMo-V2.6-Flash22.7%
artificialanalysis.ai · 2026-09-24
LiveCodeBench
third-party
~#5/20~88.7%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #4Nemotron 3 Ultra89.0%
  3. ~#5DeepSeek V4.1 FlashEstimated88.7%
  4. #5gpt-oss-120b87.8%
estimated from related benchmarks
~#3/20~90.7%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. ~#3Kimi K3Estimated90.7%
  4. #3Qwen3.8 27B90.3%
estimated from related benchmarks
#2/1991.9%Lab-reported
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #3Qwen3.8 27B90.3%
huggingface.co · 2026-08-26
SciCode
third-party
#24/9551.9%Verified
  1. #1Claude Opus 5.566.9%
  2. #22Qwen3.8 Max (0902)52.1%
  3. #24DeepSeek V4.1 Flash51.9%
  4. #25GLM-5.3 Flash51.6%
artificialanalysis.ai · 2026-09-24
#5/9559.5%BestVerified
  1. #1Claude Opus 5.566.9%
  2. #4MiMo-V2.6-Pro60.9%
  3. #5Kimi K359.5%
  4. #6GLM-5.359.0%
artificialanalysis.ai · 2026-09-24
#28/9550.6%Verified
  1. #1Claude Opus 5.566.9%
  2. #27DeepSeek V4 Pro (0813)51.0%
  3. #28Qwen3.8 Flash-Next50.6%
  4. #29Inkling Small49.7%
artificialanalysis.ai · 2026-09-24
AA Coding Agent Index
third-party
~#10/15~48.0Estimated
  1. #1Claude Opus 5.566.0
  2. #9Kimi K351.9
  3. ~#10DeepSeek V4.1 FlashEstimated48.0
  4. #10Grok 4.647.0
estimated from related benchmarks
#9/1451.9Verified
  1. #1Claude Opus 5.566.0
  2. #8GLM-5.353.6
  3. #9Kimi K351.9
  4. #10Grok 4.647.0
artificialanalysis.ai · 2026-09-24
~#10/15~47.7Estimated
  1. #1Claude Opus 5.566.0
  2. #9Kimi K351.9
  3. ~#10Qwen3.8 Flash-NextEstimated47.7
  4. #10Grok 4.647.0
estimated from related benchmarks
Not comparable across labs (1)
SWE-bench Pro (Qwen-corrected)
day 0
—
—
#3/462.5%Lab-reported
  1. #1Qwen3.8 2.4T A95B67.7%
  2. #2Qwen3.8-Omni-Flash63.3%
  3. #3Qwen3.8 Flash-Next62.5%
  4. #4Qwen3.8 27B61.7%
huggingface.co · 2026-08-26
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
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
IFBench
day 0
~#19/126~75.5%Estimated
  1. #1MiniMax-M382.9%
  2. #18Gemma 4 31B75.6%
  3. ~#19DeepSeek V4.1 FlashEstimated75.5%
  4. #19GPT-5.3 Codex75.4%
estimated from related benchmarks
~#3/126~82.6%Estimated
  1. #1MiniMax-M382.9%
  2. #2Qwen3.8 2.4T A95B82.8%
  3. ~#3Kimi K3Estimated82.6%
  4. #3Qwen3.8-Omni-Flash81.5%
estimated from related benchmarks
#5/12581.3%Lab-reported
  1. #1MiniMax-M382.9%
  2. #4Nemotron 3 Ultra81.4%
  3. #5Qwen3.8 Flash-Next81.3%
  4. #6Intern-S2-Preview (35B-A3B)80.9%
huggingface.co · 2026-08-26
MMMLU
day 0
~#2/5~90.8%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2DeepSeek V4.1 FlashEstimated90.8%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
~#2/5~92.0%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2Kimi K3Estimated92.0%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
~#2/5~91.5%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2Qwen3.8 Flash-NextEstimated91.5%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
LMArena Text, Non-English (Elo)
third-party
#12/531462Verified
  1. #1Claude Opus 5.51501
  2. #11GLM-5.3 Flash1463
  3. #12DeepSeek V4.1 Flash1462
  4. #13MiMo-V2.6-Pro1461
lmarena.ai · 2026-09-25
#7/531477BestVerified
  1. #1Claude Opus 5.51501
  2. #6Gemini 3.1 Pro1480
  3. #7Kimi K31477
  4. #8GPT-5.51474
lmarena.ai · 2026-09-25
~#15/54~1457Estimated
  1. #1Claude Opus 5.51501
  2. #14GPT-6 AstraFew votes1460
  3. ~#15Qwen3.8 Flash-NextEstimated1457
  4. #15GPT-5.6 Terra1456
estimated from related benchmarks
LMArena Text (Elo)
third-party
#13/531477Verified
  1. #1Claude Opus 5.51509
  2. #10MiMo-V2.6-Pro1480
  3. #13DeepSeek V4.1 Flash1477
  4. #14GLM-5.3 Flash1475
lmarena.ai · 2026-09-25
#6/531488BestVerified
  1. #1Claude Opus 5.51509
  2. #3Muse Spark 1.31493
  3. #6Kimi K31488
  4. #7Gemini 3.1 Pro1487
lmarena.ai · 2026-09-25
~#15/54~1468Estimated
  1. #1Claude Opus 5.51509
  2. #14GLM-5.3 Flash1475
  3. ~#15Qwen3.8 Flash-NextEstimated1468
  4. #15GPT-5.6 Terra1466
estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)
third-party
#20/251540Verified
  1. #1GPT-6 Astra2173
  2. #19Gemini 3.5 Flash-Lite1559
  3. #20DeepSeek V4.1 Flash1540
  4. #21Gemini 3.1 Pro1491
eqbench.com · 2026-09-24
#5/252082BestVerified
  1. #1GPT-6 Astra2173
  2. #4GPT-6 Sol2125
  3. #5Kimi K32082
  4. #6GLM-5.32075
eqbench.com · 2026-09-24
~#9/26~1947Estimated
  1. #1GPT-6 Astra2173
  2. #8Grok 4.72007
  3. ~#9Qwen3.8 Flash-NextEstimated1947
  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
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
MMMU
day 0
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2DeepSeek V4.1 FlashEstimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2Kimi K3Estimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2Qwen3.8 Flash-NextEstimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
MMMU-Pro
day 0
#22/7677.0%Verified
  1. #1Claude Opus 5.587.7%
  2. #20Qwen3.5 397B A17B77.3%
  3. #22DeepSeek V4.1 Flash77.0%
  4. #23Intern-S2-Preview (35B-A3B)76.9%
artificialanalysis.ai · 2026-09-24
#11/7680.5%BestVerified
  1. #1Claude Opus 5.587.7%
  2. #10GPT-5.6 Terra80.7%
  3. #11Kimi K380.5%
  4. #11Qwen3.7 Plus80.5%
artificialanalysis.ai · 2026-09-24
#14/7679.8%Verified
  1. #1Claude Opus 5.587.7%
  2. #13GPT-5.579.9%
  3. #14Qwen3.8 Flash-Next79.8%
  4. #15Apodex 1.179.2%
artificialanalysis.ai · 2026-09-24
CharXiv Reasoning
day 0
~#5/9~79.2%Estimated
  1. #1Gemini 3.8 Flash86.2%
  2. #4Qwen3.8-Omni-Flash83.5%
  3. ~#5DeepSeek V4.1 FlashEstimated79.2%
  4. #5Muse Glimmer78.8%
estimated from related benchmarks
#2/884.8%BestLab-reported
  1. #1Gemini 3.8 Flash86.2%
  2. #2Kimi K384.8%
  3. #3Qwen3.8 Flash-Next84.6%
huggingface.co · 2026-07-16
#3/884.6%Lab-reported
  1. #1Gemini 3.8 Flash86.2%
  2. #2Kimi K384.8%
  3. #3Qwen3.8 Flash-Next84.6%
  4. #4Qwen3.8-Omni-Flash83.5%
huggingface.co · 2026-08-26
OmniDocBench v1.5
third-party
~#5/8~85.2%Estimated
  1. #1MiniMax-M391.6%
  2. #4Gemini 3.1 Pro85.3%
  3. ~#5DeepSeek V4.1 FlashEstimated85.2%
  4. #5Claude Haiku 4.579.6%
estimated from related benchmarks
#2/791.1%Lab-reported
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #2Qwen3.8 27B91.1%
huggingface.co · 2026-07-16
~#4/8~88.9%Estimated
  1. #1MiniMax-M391.6%
  2. #2Qwen3.8 27B91.1%
  3. ~#4Qwen3.8 Flash-NextEstimated88.9%
  4. #4Gemini 3.1 Pro85.3%
estimated from related benchmarks
GDP.pdf (AA)
third-party
#24/4212.8%Verified
  1. #1GPT-6 Astra31.0%
  2. #23Claude Sonnet 513.2%
  3. #24DeepSeek V4.1 Flash12.8%
  4. #24Inkling12.8%
artificialanalysis.ai · 2026-09-24
#8/4222.0%BestVerified
  1. #1GPT-6 Astra31.0%
  2. #7Qwen3.8 Max (0902)22.8%
  3. #8Kimi K322.0%
  4. #9Claude Opus 521.6%
artificialanalysis.ai · 2026-09-24
#18/4215.6%Verified
  1. #1GPT-6 Astra31.0%
  2. #17Qwen3.8 27B16.6%
  3. #18Qwen3.8 Flash-Next15.6%
  4. #19GLM-5.3 Flash15.4%
artificialanalysis.ai · 2026-09-24
Chartography
third-party
#19/2211.6%Verified
  1. #1GPT-6 Astra71.0%
  2. #18Grok 4.714.7%
  3. #19DeepSeek V4.1 Flash11.6%
  4. #20Gemini 3.5 Flash-Lite10.2%
surgehq.ai · 2026-09-24
#12/2226.6%BestVerified
  1. #1GPT-6 Astra71.0%
  2. #11Claude Opus 527.3%
  3. #12Kimi K326.6%
  4. #13Gemini 3.1 Pro26.1%
surgehq.ai · 2026-09-24
~#10/23~27.9%Estimated
  1. #1GPT-6 Astra71.0%
  2. #8Qwen3.8 Max (0902)29.1%
  3. ~#10Qwen3.8 Flash-NextEstimated27.9%
  4. #10Muse Spark 1.327.6%
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
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
GPQA Diamond
day 0
#23/17290.9%Lab-reported
  1. #1GPT-6 Astra96.1%
  2. #22Qwen3.8-Omni-Flash91.0%
  3. #23DeepSeek V4.1 Flash90.9%
  4. #24Qwen3.8 27B90.5%
huggingface.co · 2026-09-10
#5/17293.5%BestVerified
  1. #1GPT-6 Astra96.1%
  2. #5Grok 4.693.5%
  3. #5Kimi K393.5%
  4. #5Muse Spark 1.393.5%
artificialanalysis.ai · 2026-09-24
#16/17292.3%Verified
  1. #1GPT-6 Astra96.1%
  2. #16Hy4 preview92.3%
  3. #16Qwen3.8 Flash-Next92.3%
  4. #18GLM-5.391.7%
artificialanalysis.ai · 2026-09-24
Humanity's Last Exam
day 0
#25/17839.2%Verified
  1. #1Claude Opus 5.561.4%
  2. #24GLM-5.3 Flash39.9%
  3. #25DeepSeek V4.1 Flash39.2%
  4. #26MiniMax-M339.0%
artificialanalysis.ai · 2026-09-24
#11/17846.9%BestVerified
  1. #1Claude Opus 5.561.4%
  2. #10Gemini 3.1 Pro47.0%
  3. #11Kimi K346.9%
  4. #12Step 5 Preview46.5%
artificialanalysis.ai · 2026-09-24
#29/17838.0%Verified
  1. #1Claude Opus 5.561.4%
  2. #28Grok Build 0.1 061638.3%
  3. #29Qwen3.8 Flash-Next38.0%
  4. #30Motif 337.0%
artificialanalysis.ai · 2026-09-24
AIME (latest)
day 0
~#2/25~96.7%Estimated
  1. #1Inkling97.1%
  2. ~#2DeepSeek V4.1 FlashEstimated96.7%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
~#2/25~96.8%Estimated
  1. #1Inkling97.1%
  2. ~#2Kimi K3Estimated96.8%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
~#2/25~96.7%Estimated
  1. #1Inkling97.1%
  2. ~#2Qwen3.8 Flash-NextEstimated96.7%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
AA Intelligence Index
third-party
#20/7439.5Verified
  1. #1Claude Opus 5.557.6
  2. #19Qwen3.8 Flash-Next39.8
  3. #20DeepSeek V4.1 Flash39.5
  4. #21GPT-5.538.4
artificialanalysis.ai · 2026-09-24
#14/7443.6BestVerified
  1. #1Claude Opus 5.557.6
  2. #13Step 5 Preview43.7
  3. #14Kimi K343.6
  4. #15GPT-5.6 Terra42.1
artificialanalysis.ai · 2026-09-24
#19/7439.8Verified
  1. #1Claude Opus 5.557.6
  2. #18Qwen3.8 2.4T A95B39.9
  3. #19Qwen3.8 Flash-Next39.8
  4. #20DeepSeek V4.1 Flash39.5
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
DeepSeek V4.1 Flash
Kimi K3
Qwen3.8 Flash-Next
τ²-bench (Telecom)
day 0
~#4/116~95.5%Estimated
  1. #1Step 3.7 Flash98.5%
  2. #2Qwen3.5 397B A17B95.6%
  3. ~#4DeepSeek V4.1 FlashEstimated95.5%
  4. #4Qwen3.6 35B A3B95.3%
estimated from related benchmarks
~#2/116~98.3%Estimated
  1. #1Step 3.7 Flash98.5%
  2. ~#2Kimi K3Estimated98.3%
  3. #2Gemini 3.1 Pro95.6%
estimated from related benchmarks
~#2/116~98.3%Estimated
  1. #1Step 3.7 Flash98.5%
  2. ~#2Qwen3.8 Flash-NextEstimated98.3%
  3. #2Gemini 3.1 Pro95.6%
estimated from related benchmarks
OSWorld-Verified
day 0
~#5/12~81.7%Estimated
  1. #1Kimi K384.8%
  2. #4MiMo-V2.6-Pro82.0%
  3. ~#5DeepSeek V4.1 FlashEstimated81.7%
  4. #5Claude Sonnet 581.2%
estimated from related benchmarks
#1/1184.8%Lab-reported
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
huggingface.co · 2026-07-16
~#3/12~83.6%Estimated
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. ~#3Qwen3.8 Flash-NextEstimated83.6%
  4. #3Nex-N2.5-Pro82.2%
estimated from related benchmarks
OSWorld 2.0
day 0
~#3/5~57.7%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3DeepSeek V4.1 FlashEstimated57.7%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
#2/458.3%Lab-reported
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. #3GPT-5.6 Terra50.2%
huggingface.co · 2026-07-16
~#3/5~56.6%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Qwen3.8 Flash-NextEstimated56.6%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
BrowseComp
day 0
~#7/18~87.6%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #6Step 5 Preview88.7%
  3. ~#7DeepSeek V4.1 FlashEstimated87.6%
  4. #7GPT-5.6 Terra87.5%
estimated from related benchmarks
#3/1791.2%Lab-reported
  1. #1Atria Dawn Preview92.5%
  2. #2GPT-6 Astra91.5%
  3. #3Kimi K391.2%
  4. #4Claude Opus 590.8%
huggingface.co · 2026-07-16
~#5/18~90.4%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #4Claude Opus 590.8%
  3. ~#5Qwen3.8 Flash-NextEstimated90.4%
  4. #5Nex-N2.5-Pro89.7%
estimated from related benchmarks
GDPval-AA (Elo)
third-party
#12/861600Verified
  1. #1Claude Opus 5.51846
  2. #11Qwen3.8 Flash-Next1612
  3. #12DeepSeek V4.1 Flash1600
  4. #13Qwen3.8 2.4T A95B1597
artificialanalysis.ai · 2026-09-24
#16/861524Verified
  1. #1Claude Opus 5.51846
  2. #15GPT-6 Astra1542
  3. #16Kimi K31524
  4. #17GPT-6 Sol1487
artificialanalysis.ai · 2026-09-24
#11/861612BestVerified
  1. #1Claude Opus 5.51846
  2. #10Grok 4.61632
  3. #11Qwen3.8 Flash-Next1612
  4. #12DeepSeek V4.1 Flash1600
artificialanalysis.ai · 2026-09-24
AutomationBench-AA
third-party
#2/7968.9%BestVerified
  1. #1Claude Opus 5.569.5%
  2. #2DeepSeek V4.1 Flash68.9%
  3. #3GPT-6 Astra68.5%
artificialanalysis.ai · 2026-09-24
#13/7958.3%Verified
  1. #1Claude Opus 5.569.5%
  2. #12MiMo-V2.6-Pro58.6%
  3. #13Kimi K358.3%
  4. #14Muse Spark 1.357.9%
artificialanalysis.ai · 2026-09-24
#19/7955.9%Verified
  1. #1Claude Opus 5.569.5%
  2. #18Qwen3.8 Max (0902)56.2%
  3. #19Qwen3.8 Flash-Next55.9%
  4. #20GPT-6 Luna53.2%
artificialanalysis.ai · 2026-09-24
τ-Bench Banking (AA)
third-party
~#14/89~40.7%Estimated
  1. #1Muse Spark 1.350.5%
  2. #13GPT-6 Astra41.4%
  3. ~#14DeepSeek V4.1 FlashEstimated40.7%
  4. #14GPT-5.6 Terra40.2%
estimated from related benchmarks
#8/8846.0%BestVerified
  1. #1Muse Spark 1.350.5%
  2. #6GLM-5.3 Flash47.2%
  3. #8Kimi K346.0%
  4. #9Qwen3.8 Flash-Next45.4%
artificialanalysis.ai · 2026-09-24
#9/8845.4%Verified
  1. #1Muse Spark 1.350.5%
  2. #8Kimi K346.0%
  3. #9Qwen3.8 Flash-Next45.4%
  4. #10Gemini 3.8 Flash44.9%
artificialanalysis.ai · 2026-09-24
Not comparable across labs (1)
OSWorld 2.0 (release not stated)
day 0
—
—
#3/452.3%Lab-reported
  1. #1Muse Spark 1.366.9%
  2. #2Nex-N2.5-Pro56.4%
  3. #3Qwen3.8 Flash-Next52.3%
  4. #4Nex-N2.5-Mini30.5%
huggingface.co · 2026-08-26