Skip to content
Model Pareto

Gemini 4 Carbon vs Muse Wonder Full

In hard reasoning, Gemini 4 Carbon scores 99.7 and Muse Wonder Full 93.4 on the 0–100 quality scale. Muse Wonder Full has no listed API price; Gemini 4 Carbon costs $8.00 per million tokens, blended 3:1 input:output.
Gemini 4 CarbonMuse Wonder Full
Focus

Hard reasoning: quality vs price

Compared models are ringed; the other 213 ranked here are greyed.

Lower price is better. Pareto frontier: Gemma 4 E4B, Sarvam 30B, Qwen3.5 4B, Sarvam 105B, gpt-oss-20b, Ling-3.0-flash-VL, MiMo-V2.5, Ternary Bonsai 2 27B, Claude Haiku 5.5, MiniMax-M3.1-Flash, MiMo-V2.6-Pro, GPT-6.1 Terra, Claude Sonnet 5.5, Gemini 4 Carbon. 66 models are not plotted: Muse Wonder Full, Muse Spark Open, Beam 1.0 501B A23B, DeepSeek V4.1 Flash NVFP4, GLM-5.3 Flash NVFP4, GLM-5.3 NVFP4, Qwen3.8 Flash-Next NVFP4, DeepSeek-V4-Pro-0813-nvfp4-DSpark, Kimi-K3-INT4, Qwen3.8 27B NVFP4, Motif 3, Nex-N2-Pro, Ternary Bonsai 27B, Nemotron 3.5 Super VL NVFP4, GLM-5.3-MXFP4, K2 Horizon 375B A23B, Qwen3.8 27B MXFP4, Xing4.0-29B-A4B, Nemotron 3.5 Super VL, Solar Open2 250B, Bonsai 27B, A.X-K2, Nemotron 3.5 Bolt 74B-A9B, K2 Horizon MoVA 36B A4B, K2-Horizon-32B, Maple-Preview, Intern-S2-Preview (35B-A3B), K-EXAONE 2.0 0803, G9v3-39A5B, Phi-4-reasoning-plus, K2 Horizon 7B, EXAONE 4.5 33B, Nanbeige4.1-3B, HyperNova 60B 2605, Nemotron Cascade 2 30B A3B, INTELLECT-3, Apriel-v1.6-15B-Thinker, K2 Horizon 3.7B, NVIDIA Nemotron-Labs Dory1.0 4B, LFM2.5-24B-A2B, DiffusionGemma 26B A4B, North Mini Code, Falcon-H1R-7B, Ling 3.0 Tiny, MiniCPM5-2B, LongCat Flash Lite, HyperCLOVA X SEED Think (32B), Tri-21B-Think, Olmo 3.1 32B Think, LFM2.5-2.6B, LFM2.5-8B-A1B, Olmo 3.1 32B Instruct, NVIDIA Nemotron 3 Nano 4B, Phi-4-mini-flash-reasoning, Falcon-H1-34B-Instruct, Qwen3.5 2B, Hermes 4 - Llama-3.1 70B, LFM2 24B A2B, Gemma 4 E2B, MiniCPM5-1B, Jamba 1.7 Large, MiniCPM-V 4.6 1.3B, Jamba Reasoning 3B, Tiny Aya Global, Qwen3.5 0.8B, K2 Horizon 0.9B.

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

Muse Wonder Full, Muse Spark Open, Beam 1.0 501B A23B, DeepSeek V4.1 Flash NVFP4, GLM-5.3 Flash NVFP4, GLM-5.3 NVFP4, Qwen3.8 Flash-Next NVFP4, DeepSeek-V4-Pro-0813-nvfp4-DSpark, Kimi-K3-INT4, Qwen3.8 27B NVFP4, Motif 3, Nex-N2-Pro, Ternary Bonsai 27B, Nemotron 3.5 Super VL NVFP4, GLM-5.3-MXFP4, K2 Horizon 375B A23B, Qwen3.8 27B MXFP4, Xing4.0-29B-A4B, Nemotron 3.5 Super VL, Solar Open2 250B, Bonsai 27B, A.X-K2, Nemotron 3.5 Bolt 74B-A9B, K2 Horizon MoVA 36B A4B, K2-Horizon-32B, Maple-Preview, Intern-S2-Preview (35B-A3B), K-EXAONE 2.0 0803, G9v3-39A5B, Phi-4-reasoning-plus, K2 Horizon 7B, EXAONE 4.5 33B, Nanbeige4.1-3B, HyperNova 60B 2605, Nemotron Cascade 2 30B A3B, INTELLECT-3, Apriel-v1.6-15B-Thinker, K2 Horizon 3.7B, NVIDIA Nemotron-Labs Dory1.0 4B, LFM2.5-24B-A2B, DiffusionGemma 26B A4B, North Mini Code, Falcon-H1R-7B, Ling 3.0 Tiny, MiniCPM5-2B, LongCat Flash Lite, HyperCLOVA X SEED Think (32B), Tri-21B-Think, Olmo 3.1 32B Think, LFM2.5-2.6B, LFM2.5-8B-A1B, Olmo 3.1 32B Instruct, NVIDIA Nemotron 3 Nano 4B, Phi-4-mini-flash-reasoning, Falcon-H1-34B-Instruct, Qwen3.5 2B, Hermes 4 - Llama-3.1 70B, LFM2 24B A2B, Gemma 4 E2B, MiniCPM5-1B, Jamba 1.7 Large, MiniCPM-V 4.6 1.3B, Jamba Reasoning 3B, Tiny Aya Global, Qwen3.5 0.8B, K2 Horizon 0.9B

  • Best-value frontier (nothing is both cheaper and better)
  • Evidencestrong → weak
  • Estimated from other categories
  • Gemini 4 Carbon
    Quality
    99.7
    Rank
    #1/215
    Price
    $8.00/M tok
    Speed
    83tok/s
    In$4.00Out$20/M tok

    vs Muse Wonder Full: +6.3 quality · price n/a

  • Muse Wonder Full
    Quality
    93.4
    Rank
    #6/215
    Price
    —
    Speed
    —

    vs Gemini 4 Carbon: −6.3 quality · price n/a

Overview

Metric
Gemini 4 Carbon
Muse Wonder Full
Lab
Google DeepMind
Meta
Released
Oct 28, 2026
Oct 12, 2026
Weights
Proprietary
Proprietary
Input price
$/M tok
$4.00
—
Output price
$/M tok
$20
—
Blended price
$/M tok · 3:1 in:out
$8.00
—
Output speed
tok/s
83
—
Context
tokens
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
Gemini 4 Carbon
Muse Wonder Full
Coding
Not ranked
91.7±4.5
#4/134Verified
Vision
Not ranked
94.4±7.5
#2/97Verified
Hard reasoning
99.7±4.8Best
#1/215Verifiedbest value
93.4±5.1
#6/215Verified
Agents
Not ranked
99.0±4.4
#2/186Verified

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
Gemini 4 Carbon
Muse Wonder Full
GPQA Diamond
day 0
~#2/191~95.9%Estimated
  1. #1GPT-6 Astra96.1%
  2. ~#2Gemini 4 CarbonEstimated95.9%
  3. #2Gemini 3.8 Flash95.3%
estimated from related benchmarks
~#2/191~96.1%Estimated
  1. #1GPT-6 Astra96.1%
  2. ~#2Muse Wonder FullEstimated96.1%
  3. #2Gemini 3.8 Flash95.3%
estimated from related benchmarks
Humanity's Last Exam
day 0
~#2/198~61.2%Estimated
  1. #1Claude Opus 5.561.4%
  2. ~#2Gemini 4 CarbonEstimated61.2%
  3. #2Claude Fable 5.159.1%
estimated from related benchmarks
#7/19753.6%Verified
  1. #1Claude Opus 5.561.4%
  2. #6GPT-6 Astra54.7%
  3. #7Muse Wonder Full53.6%
  4. #8GPT-6.1 Sol52.9%
artificialanalysis.ai · 2026-10-13
AIME (latest)
day 0
~#2/33~98.4%Estimated
  1. #1GLM-5.3 NVFP499.5%
  2. ~#2Gemini 4 CarbonEstimated98.4%
  3. #2Beam 1.0 501B A23B97.7%
estimated from related benchmarks
~#2/33~99.1%Estimated
  1. #1GLM-5.3 NVFP499.5%
  2. ~#2Muse Wonder FullEstimated99.1%
  3. #2Beam 1.0 501B A23B97.7%
estimated from related benchmarks
AA Intelligence Index
third-party
#1/10259.4BestVerified
  1. #1Gemini 4 Carbon59.4
  2. #2Claude Opus 5.557.6
artificialanalysis.ai · 2026-10-30
#4/10255.8Verified
  1. #1Gemini 4 Carbon59.4
  2. #3Claude Sonnet 5.556.0
  3. #4Muse Wonder Full55.8
  4. #5Kimi K3.1 Pro55.1
artificialanalysis.ai · 2026-10-13
Benchmarks in other categories (3)

Coding benchmarks

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

Metric
Gemini 4 Carbon
Muse Wonder Full
SWE-bench Verified
day 0
—
~#2/20~93.8%Estimated
  1. #1Claude Opus 596.0%
  2. ~#2Muse Wonder FullEstimated93.8%
  3. #2Claude Sonnet 585.2%
estimated from related benchmarks
SWE-bench Pro (Public)
day 0
—
~#5/23~78.7%Estimated
  1. #1Claude Opus 5.589.9%
  2. #4Claude Opus 579.2%
  3. ~#5Muse Wonder FullEstimated78.7%
  4. #5Intern-S2-397B68.5%
estimated from related benchmarks
DeepSWE v1.1
day 0
—
#2/4277.0%Verified
  1. #1Gemini 4 Argon78.8%
  2. #2DeepSeek V4.1 Flash 101777.0%
  3. #2Muse Wonder Full77.0%
  4. #4Muse Wonder75.8%
artificialanalysis.ai · 2026-10-15
Terminal-Bench 4.0
day 0
—
#5/9957.1%Verified
  1. #1Kimi K3.1 Pro67.2%
  2. #5Gemini 4 Argon57.1%
  3. #5Muse Wonder Full57.1%
  4. #7GPT-6.1 Sol56.1%
artificialanalysis.ai · 2026-10-13
LiveCodeBench
third-party
—
~#2/28~92.3%Estimated
  1. #1Qwen3.8-Omni-Flash92.6%
  2. ~#2Muse Wonder FullEstimated92.3%
  3. #2Qwen3.8 Flash-Next91.9%
estimated from related benchmarks
SciCode
third-party
—
#3/11762.6%Verified
  1. #1Claude Opus 5.566.9%
  2. #2Claude Fable 5.163.1%
  3. #3Muse Wonder Full62.6%
  4. #4Gemini 4 Argon61.8%
artificialanalysis.ai · 2026-10-13
AA Coding Agent Index
third-party
—
#4/2558.9Verified
  1. #1Claude Opus 5.561.7
  2. #3Claude Sonnet 5.559.3
  3. #4Muse Wonder Full58.9
  4. #5Claude Fable 5.157.6
artificialanalysis.ai · 2026-10-24

Vision benchmarks

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

Metric
Gemini 4 Carbon
Muse Wonder Full
MMMU
day 0
—
~#2/3~71.3%Estimated
  1. #1Claude Haiku 4.573.2%
  2. ~#2Muse Wonder FullEstimated71.3%
  3. #2Llama 4 Maverick69.4%
estimated from related benchmarks
MMMU-Pro
day 0
—
~#2/88~87.6%Estimated
  1. #1Claude Opus 5.587.7%
  2. ~#2Muse Wonder FullEstimated87.6%
  3. #2GPT-6 Astra86.9%
estimated from related benchmarks
CharXiv Reasoning
day 0
—
~#2/12~86.0%Estimated
  1. #1Gemini 3.8 Flash86.2%
  2. ~#2Muse Wonder FullEstimated86.0%
  3. #2Kimi K384.8%
estimated from related benchmarks
OmniDocBench v1.5
third-party
—
~#2/8~91.1%Estimated
  1. #1MiniMax-M391.6%
  2. ~#2Muse Wonder FullEstimated91.1%
  3. #2Kimi K391.1%
estimated from related benchmarks
GDP.pdf (AA)
third-party
—
#1/4632.6%Verified
  1. #1Muse Wonder Full32.6%
  2. #2GPT-6 Astra31.0%
artificialanalysis.ai · 2026-10-13
Chartography
third-party
—
~#4/24~53.5%Estimated
  1. #1GPT-6 Astra71.0%
  2. #3GPT-6 Sol53.6%
  3. ~#4Muse Wonder FullEstimated53.5%
  4. #4Claude Haiku 5.546.4%
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
Gemini 4 Carbon
Muse Wonder Full
τ²-bench (Telecom)
day 0
—
~#2/120~98.3%Estimated
  1. #1Step 3.7 Flash98.5%
  2. ~#2Muse Wonder FullEstimated98.3%
  3. #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
estimated from related benchmarks
OSWorld-Verified
day 0
—
~#2/15~84.5%Estimated
  1. #1Kimi K384.8%
  2. ~#2Muse Wonder FullEstimated84.5%
  3. #2Qwen3.8 27B84.3%
estimated from related benchmarks
OSWorld 2.0
day 0
—
~#2/5~58.6%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. ~#2Muse Wonder FullEstimated58.6%
  3. #2Kimi K358.3%
estimated from related benchmarks
BrowseComp
day 0
—
~#2/22~92.2%Estimated
  1. #1Atria Dawn Preview92.5%
  2. ~#2Muse Wonder FullEstimated92.2%
  3. #2GPT-6 Astra91.5%
estimated from related benchmarks
GDPval-AA (Elo)
third-party
—
#2/901857Verified
  1. #1Claude Opus 5.51866
  2. #2Muse Wonder Full1857
  3. #3Muse Wonder1770
artificialanalysis.ai · 2026-10-13
AutomationBench-AA
third-party
—
#1/8371.3%Verified
  1. #1Muse Wonder Full71.3%
  2. #2Claude Opus 5.569.5%
artificialanalysis.ai · 2026-10-13
τ-Bench Banking (AA)
third-party
—
~#3/91~49.9%Estimated
  1. #1Muse Spark 1.350.5%
  2. #2GLM-5.350.3%
  3. ~#3Muse Wonder FullEstimated49.9%
  4. #3Qwen3.8 2.4T A95B49.1%
estimated from related benchmarks