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.
Hy3Intern-S2-Preview (35B-A3B)Xing4.0-29B-A4B
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
  • Hy3
    Quality
    34.4
    Rank
    #61/103
    Price
    $0.25/M tok
    Speed
    86tok/s
    In$0.14Out$0.58/M tok

    vs Xing4.0-29B-A4B: −27.8 quality · price n/a

  • Intern-S2-Preview (35B-A3B)
    Quality
    30.2
    Rank
    #73/103
    Price
    —
    Speed
    —

    vs Xing4.0-29B-A4B: −31.9 quality · price n/a

  • Xing4.0-29B-A4B
    Quality
    62.1
    Rank
    #19/103
    Price
    —
    Speed
    —

    vs Hy3: +27.8 quality · price n/a

Overview

Metric
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
Lab
Tencent Hunyuan
Shanghai AI Lab (Intern)
China Telecom AI (TeleAI)
Released
Jul 6, 2026
May 15, 2026
Sep 17, 2026
Weights
Open
Open
Open
Input price
$/M tok
$0.14
—
—
Output price
$/M tok
$0.58
—
—
Blended price
$/M tok · 3:1 in:out
$0.25
—
—
Output speed
tok/s
86
—
—
Context
tokens
256K
262KBest
262KBest

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
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
Coding
34.4±4.3
#61/103Mixed
30.2±8.6
#73/103Lab-reported
62.1±10.6Best
#19/103Lab-reported
Writing & chat
67.7±6.0Best
#30/146Verified
67.7±9.7
#31/146Lab-reported
67.4±9.7
#32/146Lab-reported
Vision
Not ranked
64.3±8.7
#22/82Lab-reported
Not ranked
Hard reasoning
64.1±3.7Best
#41/182Verified
53.6±8.6
#59/182Lab-reported
59.8±10.0
#46/182Lab-reported
Agents
45.1±4.6
#48/173Mixed
Not ranked
Not ranked

Coding benchmarks

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

Metric
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
SWE-bench Verified
day 0
#5/1378.0%BestLab-reported
  1. #1Claude Opus 596.0%
  2. #4MiniMax-M380.5%
  3. #5Hy378.0%
  4. #6Inkling77.6%
huggingface.co · 2026-07-06
#12/1364.0%Lab-reported
  1. #1Claude Opus 596.0%
  2. #11Nemotron 3 Ultra70.7%
  3. #12Intern-S2-Preview (35B-A3B)64.0%
  4. #13gpt-oss-120b62.4%
huggingface.co · 2026-05-15
#9/1375.0%Lab-reported
  1. #1Claude Opus 596.0%
  2. #8Muse Glimmer76.0%
  3. #9Xing4.0-29B-A4B75.0%
  4. #10Claude Haiku 4.573.3%
huggingface.co · 2026-09-17
SWE-bench Pro (Public)
day 0
#13/1857.9%Lab-reported
  1. #1Claude Opus 5.589.9%
  2. #12GPT-5.558.6%
  3. #13Hy357.9%
  4. #14Inkling54.3%
huggingface.co · 2026-07-06
~#18/19~46.0%Estimated
  1. #1Claude Opus 5.589.9%
  2. #17Muse Glimmer51.2%
  3. ~#18Intern-S2-Preview (35B-A3B)Estimated46.0%
  4. #18Nex-N2.5-Mini43.8%
estimated from related benchmarks
~#14/19~55.3%Estimated
  1. #1Claude Opus 5.589.9%
  2. #13Hy357.9%
  3. ~#14Xing4.0-29B-A4BEstimated55.3%
  4. #14Inkling54.3%
estimated from related benchmarks
DeepSWE v1.1
day 0
#30/3028.0%Lab-reported
  1. #1DeepSeek V4.1 Flash74.2%
  2. #29Nex-N2.5-Mini36.1%
  3. #30Hy328.0%
huggingface.co · 2026-07-06
~#28/31~44.1%Estimated
  1. #1DeepSeek V4.1 Flash74.2%
  2. #27Qwen3.8 Max (0902)51.0%
  3. ~#28Intern-S2-Preview (35B-A3B)Estimated44.1%
  4. #28Qwen3.8 27B42.2%
estimated from related benchmarks
~#18/31~63.9%Estimated
  1. #1DeepSeek V4.1 Flash74.2%
  2. #16Hy4 preview64.3%
  3. ~#18Xing4.0-29B-A4BEstimated63.9%
  4. #18GPT-6 Luna63.7%
estimated from related benchmarks
Terminal-Bench 4.0
day 0
#39/830.5%Verified
  1. #1Claude Sonnet 5.563.6%
  2. #39GPT-5.4 nano0.5%
  3. #39Hy30.5%
  4. #39Muse Glimmer0.5%
artificialanalysis.ai · 2026-09-24
~#27/84~7.5%Estimated
  1. #1Claude Sonnet 5.563.6%
  2. #26Qwen3.8 2.4T A95B11.1%
  3. ~#27Intern-S2-Preview (35B-A3B)Estimated7.5%
  4. #27Qwen3.8 27B5.6%
estimated from related benchmarks
~#18/84~21.3%Estimated
  1. #1Claude Sonnet 5.563.6%
  2. #17MiMo-V2.6-Flash22.7%
  3. ~#18Xing4.0-29B-A4BEstimated21.3%
  4. #18Gemini 3.8 Flash19.7%
estimated from related benchmarks
LiveCodeBench
third-party
~#7/20~78.0%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #6Gemma 4 31B80.0%
  3. ~#7Hy3Estimated78.0%
  4. #7gpt-oss-20b77.7%
estimated from related benchmarks
~#8/20~70.5%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #7gpt-oss-20b77.7%
  3. ~#8Intern-S2-Preview (35B-A3B)Estimated70.5%
  4. #8Mistral Small 464.0%
estimated from related benchmarks
~#6/20~84.6%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #5gpt-oss-120b87.8%
  3. ~#6Xing4.0-29B-A4BEstimated84.6%
  4. #6Gemma 4 31B80.0%
estimated from related benchmarks
SciCode
third-party
#30/9548.6%BestVerified
  1. #1Claude Opus 5.566.9%
  2. #29Inkling Small49.7%
  3. #30Hy348.6%
  4. #31Quasar 438B48.1%
artificialanalysis.ai · 2026-09-24
#64/9539.6%Lab-reported
  1. #1Claude Opus 5.566.9%
  2. #63Qwen3.5 122B A10B39.7%
  3. #64Intern-S2-Preview (35B-A3B)39.6%
  4. #65gpt-oss-20b38.9%
huggingface.co · 2026-05-15
~#16/96~55.7%Estimated
  1. #1Claude Opus 5.566.9%
  2. #15GPT-5.555.8%
  3. ~#16Xing4.0-29B-A4BEstimated55.7%
  4. #16GPT-5.6 Terra55.0%
estimated from related benchmarks
AA Coding Agent Index
third-party
~last/15below measured range (<41.1)Estimated
  1. #1Claude Opus 5.566.0
  2. #14GPT-6 Luna41.1
  3. ~#15Hy3Estimated<41.1
estimated from related benchmarks
~last/15below measured range (<41.1)Estimated
  1. #1Claude Opus 5.566.0
  2. #14GPT-6 Luna41.1
  3. ~#15Intern-S2-Preview (35B-A3B)Estimated<41.1
estimated from related benchmarks
~#5/15~57.8Estimated
  1. #1Claude Opus 5.566.0
  2. #4Claude Opus 559.7
  3. ~#5Xing4.0-29B-A4BEstimated57.8
  4. #5GPT-6 Sol56.7
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
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
IFBench
day 0
~#33/126~67.2%Estimated
  1. #1MiniMax-M382.9%
  2. #32Step 3.7 Flash67.3%
  3. ~#33Hy3Estimated67.2%
  4. #33MiMo-V2.567.1%
estimated from related benchmarks
#6/12580.9%BestLab-reported
  1. #1MiniMax-M382.9%
  2. #5Qwen3.8 Flash-Next81.3%
  3. #6Intern-S2-Preview (35B-A3B)80.9%
  4. #7Nemotron Cascade 2 30B A3B80.4%
huggingface.co · 2026-05-15
#28/12569.7%Lab-reported
  1. #1MiniMax-M382.9%
  2. #27Nova 2.0 Lite70.7%
  3. #28Xing4.0-29B-A4B69.7%
  4. #29Apriel-v1.6-15B-Thinker69.1%
huggingface.co · 2026-09-17
MMMLU
day 0
~#3/5~87.8%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. #2Gemma 4 31B88.4%
  3. ~#3Hy3Estimated87.8%
  4. #3Claude Haiku 4.583.0%
estimated from related benchmarks
~#2/5~89.1%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2Intern-S2-Preview (35B-A3B)Estimated89.1%
  3. #2Gemma 4 31B88.4%
estimated from related benchmarks
~#3/5~87.7%Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. #2Gemma 4 31B88.4%
  3. ~#3Xing4.0-29B-A4BEstimated87.7%
  4. #3Claude Haiku 4.583.0%
estimated from related benchmarks
LMArena Text, Non-English (Elo)
third-party
#22/531440Verified
  1. #1Claude Opus 5.51501
  2. #21Qwen3.7 Plus1443
  3. #22Hy31440
  4. #23GPT-5.4 mini1439
lmarena.ai · 2026-09-24
~#32/54~1420Estimated
  1. #1Claude Opus 5.51501
  2. #31Gemma 4 26B A4BFew votes1423
  3. ~#32Intern-S2-Preview (35B-A3B)Estimated1420
  4. #32Muse GlimmerFew votes1418
estimated from related benchmarks
~#30/54~1429Estimated
  1. #1Claude Opus 5.51501
  2. #29Qwen3.8 27B1429
  3. ~#30Xing4.0-29B-A4BEstimated1429
  4. #30MiniMax-M31427
estimated from related benchmarks
LMArena Text (Elo)
third-party
#19/531456Verified
  1. #1Claude Opus 5.51509
  2. #19Gemini 3.5 Flash-Lite1456
  3. #19Hy31456
  4. #19Qwen3.7 Plus1456
lmarena.ai · 2026-09-24
~#33/54~1434Estimated
  1. #1Claude Opus 5.51509
  2. #32MiMo-V2.51434
  3. ~#33Intern-S2-Preview (35B-A3B)Estimated1434
  4. #33Mistral Medium 3.51426
estimated from related benchmarks
~#27/54~1442Estimated
  1. #1Claude Opus 5.51509
  2. #25Qwen3.5 397B A17B1442
  3. ~#27Xing4.0-29B-A4BEstimated1442
  4. #27MiniMax-M31441
estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)
third-party
~#13/26~1832Estimated
  1. #1GPT-6 Astra2173
  2. #12Qwen3.8 2.4T A95B1843
  3. ~#13Hy3Estimated1832
  4. #13Muse Glimmer1798
estimated from related benchmarks
~#15/26~1792Estimated
  1. #1GPT-6 Astra2173
  2. #14Claude Sonnet 51794
  3. ~#15Intern-S2-Preview (35B-A3B)Estimated1792
  4. #15Gemini 3.8 Flash1748
estimated from related benchmarks
~#9/26~1911Estimated
  1. #1GPT-6 Astra2173
  2. #8Grok 4.72007
  3. ~#9Xing4.0-29B-A4BEstimated1911
  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
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
MMMU
day 0
—
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2Intern-S2-Preview (35B-A3B)Estimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
—
MMMU-Pro
day 0
—
#23/7676.9%Lab-reported
  1. #1Claude Opus 5.587.7%
  2. #22DeepSeek V4.1 Flash77.0%
  3. #23Intern-S2-Preview (35B-A3B)76.9%
  4. #24Grok Build 0.1 061676.5%
huggingface.co · 2026-05-15
—
CharXiv Reasoning
day 0
—
~#5/9~79.5%Estimated
  1. #1Gemini 3.8 Flash86.2%
  2. #4Qwen3.8-Omni-Flash83.5%
  3. ~#5Intern-S2-Preview (35B-A3B)Estimated79.5%
  4. #5Muse Glimmer78.8%
estimated from related benchmarks
—
OmniDocBench v1.5
third-party
—
~#4/8~85.8%Estimated
  1. #1MiniMax-M391.6%
  2. #2Qwen3.8 27B91.1%
  3. ~#4Intern-S2-Preview (35B-A3B)Estimated85.8%
  4. #4Gemini 3.1 Pro85.3%
estimated from related benchmarks
—
GDP.pdf (AA)
third-party
—
~#21/43~14.5%Estimated
  1. #1GPT-6 Astra31.0%
  2. #20Step 5 Preview14.8%
  3. ~#21Intern-S2-Preview (35B-A3B)Estimated14.5%
  4. #21Gemini 3.5 Flash-Lite13.6%
estimated from related benchmarks
—
Chartography
third-party
—
~#14/23~24.4%Estimated
  1. #1GPT-6 Astra71.0%
  2. #13Gemini 3.1 Pro26.1%
  3. ~#14Intern-S2-Preview (35B-A3B)Estimated24.4%
  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
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
GPQA Diamond
day 0
#26/17289.7%Verified
  1. #1GPT-6 Astra96.1%
  2. #25Qwen3.7 Plus90.0%
  3. #26Hy389.7%
  4. #27Kimi K2.7 Code89.6%
artificialanalysis.ai · 2026-09-24
~#74/173~76.0%Estimated
  1. #1GPT-6 Astra96.1%
  2. #73Nova 2.0 Omni76.0%
  3. ~#74Intern-S2-Preview (35B-A3B)Estimated76.0%
  4. #74Nemotron Cascade 2 30B A3B75.8%
estimated from related benchmarks
~#64/173~79.4%Estimated
  1. #1GPT-6 Astra96.1%
  2. #63Nemotron 3 Super 120B A12B80.0%
  3. ~#64Xing4.0-29B-A4BEstimated79.4%
  4. #64EXAONE 4.5 33B79.4%
estimated from related benchmarks
Humanity's Last Exam
day 0
#38/17833.5%BestVerified
  1. #1Claude Opus 5.561.4%
  2. #36Nex-N2-Pro33.7%
  3. #38Hy333.5%
  4. #39Inkling Small33.3%
artificialanalysis.ai · 2026-09-24
#58/17821.9%Lab-reported
  1. #1Claude Opus 5.561.4%
  2. #56Muse Glimmer22.0%
  3. #58Intern-S2-Preview (35B-A3B)21.9%
  4. #59Ring-2.6-1T21.6%
huggingface.co · 2026-05-15
~#53/179~23.5%Estimated
  1. #1Claude Opus 5.561.4%
  2. #52Gemma 4 31B23.6%
  3. ~#53Xing4.0-29B-A4BEstimated23.5%
  4. #53K2 Horizon MoVA 36B A4B23.4%
estimated from related benchmarks
AIME (latest)
day 0
~#5/25~92.8%Estimated
  1. #1Inkling97.1%
  2. #4gpt-oss-120b93.4%
  3. ~#5Hy3Estimated92.8%
  4. #5Xing4.0-29B-A4B90.0%
estimated from related benchmarks
~#2/25~94.8%Estimated
  1. #1Inkling97.1%
  2. ~#2Intern-S2-Preview (35B-A3B)Estimated94.8%
  3. #2Muse Glimmer94.7%
estimated from related benchmarks
#5/2490.0%Lab-reported
  1. #1Inkling97.1%
  2. #4gpt-oss-120b93.4%
  3. #5Xing4.0-29B-A4B90.0%
  4. #6gpt-oss-20b89.3%
huggingface.co · 2026-09-17
AA Intelligence Index
third-party
#33/7425.3Verified
  1. #1Claude Opus 5.557.6
  2. #32Kimi K2.7 Code25.8
  3. #33Hy325.3
  4. #33K2 Horizon MoVA 36B A4B25.3
artificialanalysis.ai · 2026-09-24
~#37/75~24.7Estimated
  1. #1Claude Opus 5.557.6
  2. #36Inkling25.0
  3. ~#37Intern-S2-Preview (35B-A3B)Estimated24.7
  4. #37Ling-3.0-flash-VL24.6
estimated from related benchmarks
~#29/75~27.0Estimated
  1. #1Claude Opus 5.557.6
  2. #28MiniMax-M329.2
  3. ~#29Xing4.0-29B-A4BEstimated27.0
  4. #29Quasar 438B26.7
estimated from related benchmarks

Agents benchmarks

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

Metric
Hy3
Intern-S2-Preview (35B-A3B)
Xing4.0-29B-A4B
τ²-bench (Telecom)
day 0
~#36/116~77.2%Estimated
  1. #1Step 3.7 Flash98.5%
  2. #35EXAONE 4.5 33B78.1%
  3. ~#36Hy3Estimated77.2%
  4. #36GPT-5.4 nano76.0%
estimated from related benchmarks
—
—
OSWorld-Verified
day 0
~#10/12~70.1%Estimated
  1. #1Kimi K384.8%
  2. #9Nex-N2.5-Mini71.2%
  3. ~#10Hy3Estimated70.1%
  4. #10Muse Glimmer65.9%
estimated from related benchmarks
—
—
OSWorld 2.0
day 0
~last/5below measured range (<47.5%)Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #4GPT-5.547.5%
  3. ~#5Hy3Estimated<47.5%
estimated from related benchmarks
—
—
BrowseComp
day 0
#11/1784.2%Lab-reported
  1. #1Atria Dawn Preview92.5%
  2. #10GPT-5.584.4%
  3. #11Hy384.2%
  4. #12MiniMax-M383.5%
huggingface.co · 2026-07-06
—
—
GDPval-AA (Elo)
third-party
#34/861045Verified
  1. #1Claude Opus 5.51846
  2. #33Grok Build 0.1 06161052
  3. #34Hy31045
  4. #35Kimi K2.7 Code1025
artificialanalysis.ai · 2026-09-24
—
—
AutomationBench-AA
third-party
#36/7918.4%Verified
  1. #1Claude Opus 5.569.5%
  2. #35K2 Horizon 7B20.9%
  3. #36Hy318.4%
  4. #36MiMo-V2.518.4%
artificialanalysis.ai · 2026-09-24
—
—
τ-Bench Banking (AA)
third-party
#31/8822.9%Verified
  1. #1Muse Spark 1.350.5%
  2. #30Solar Pro 423.3%
  3. #31Hy322.9%
  4. #32G9v3-39A5B22.1%
artificialanalysis.ai · 2026-09-24
—
—