Sign Up

Decision support

Claude Sonnet 5.5 vs Gemini 4 Carbon

Compare canonical objects directly. The useful dimensions change with the object type, so model and hardware comparisons stay separate.

ModelsHardware
Claude Sonnet 5.5

Anthropic

Sources:models.devOpenRouter

Gemini 4 Carbon

Google DeepMind

Sources:models.devOpenRouter

Benchmark results

Head to head on 9 benchmarks both models report · 3 ties

  • Claude Sonnet 5.50leads
  • Gemini 4 Carbon6leads

  • Arena Text

    Arena rating · Higher is better

    • Claude Sonnet 5.5
      1472
      Arena #46CI 1466 to 1479
    • Gemini 4 Carbon
      Best1526
      +53 ahead · Arena #1CI 1516 to 1536
    Runs and sources
    Claude Sonnet 5.5
    • xhigh147295% CI 1466 to 1479 · 13,821 votes
    Gemini 4 Carbon
    • high152695% CI 1516 to 1536 · 4,282 votes
    Source:Arena Text leaderboard, 2026-11-09

  • Arena WebDev

    Arena rating · Higher is better

    Tie, within 95% CI
    • Claude Sonnet 5.5
      1773
      Arena #3CI 1763 to 1782
    • Gemini 4 Carbon
      1768
      Arena #4CI 1753 to 1783
    Runs and sources
    Claude Sonnet 5.5
    • high171595% CI 1707 to 1724 · 7,881 votes
    • xhigh177395% CI 1763 to 1782 · 5,818 votes
    Gemini 4 Carbon
    • high176895% CI 1753 to 1783 · 2,120 votes
    Source:Arena WebDev leaderboard, 2026-11-10

  • Arena Vision

    Arena rating · Higher is better

    • Claude Sonnet 5.5
      1263
      Arena #43CI 1253 to 1273
    • Gemini 4 Carbon
      Not reported
    Runs and sources
    Claude Sonnet 5.5
    • xhigh126395% CI 1253 to 1273 · 4,257 votes
    Gemini 4 Carbon

    Not reported

    Source:Arena Vision leaderboard, 2026-11-09

  • Agent Arena

    IPS score · Higher is better

    Tie, within 95% CI
    • Claude Sonnet 5.5
      +0.160
      Arena #4CI +0.131 to +0.188
    • Gemini 4 Carbon
      +0.170
      Arena #3CI +0.163 to +0.177
    Runs and sources
    Claude Sonnet 5.5
    • max+0.16095% CI +0.131 to +0.188 · 1,274,623 observations
    Gemini 4 Carbon
    • high+0.17095% CI +0.163 to +0.177 · 9,558,978 observations
    Source:Agent Arena leaderboard, 2026-11-11

  • ARC-AGI-2

    Best published score · Higher is better

    • Claude Sonnet 5.5
      83.9%
      #17 of 85
    • Gemini 4 Carbon
      Best94.2%
      +10.3 pts ahead · #2 of 85
    Runs, sources and effort chart
    Claude Sonnet 5.5
    • low24.0%
    • medium32.8%
    • high67.4%
    • xhigh83.9%
    Gemini 4 Carbon
    • low49.6%
    • medium94.2%
    • high92.1%
    Source:External evaluation via Epoch AIOriginal-source terms apply

  • FrontierMath-Tier-4-v2-Private

    Best published score · Higher is better

    • Claude Sonnet 5.5
      80.5%
      #11 of 72
    • Gemini 4 Carbon
      Best95.1%
      +14.6 pts ahead · #4 of 72
    Runs and sources
    Claude Sonnet 5.5
    • max80.5%
    Gemini 4 Carbon
    • high95.1%
    Source:Epoch AI evaluation

  • FrontierMath-Tiers-1-3-v2-Private

    Best published score · Higher is better

    • Claude Sonnet 5.5
      88.8%
      #10 of 87
    • Gemini 4 Carbon
      Best90.5%
      +1.8 pts ahead · #6 of 87
    Runs and sources
    Claude Sonnet 5.5
    • max88.8%
    Gemini 4 Carbon
    • high90.5%
    Source:Epoch AI evaluation

  • GPQA diamond

    Best published score · Higher is better

    • Claude Sonnet 5.5
      95.6%
      #4 of 161
    • Gemini 4 Carbon
      Best96.5%
      +0.9 pts ahead · #1 of 161
    Runs and sources
    Claude Sonnet 5.5
    • max95.6%
    Gemini 4 Carbon
    • high96.5%
    Source:Epoch AI evaluation

  • Humanity's Last Exam

    Best published score · Higher is better

    • Claude Sonnet 5.5
      46.1%
      #4 of 34
    • Gemini 4 Carbon
      Not reported
    Runs and sources
    Claude Sonnet 5.5
    • xhigh46.1%
    Gemini 4 Carbon

    Not reported

    Source:External evaluation via Epoch AIOriginal-source terms apply

  • OTIS Mock AIME 2024-2025

    Best published score · Higher is better

    Tie
    • Claude Sonnet 5.5
      100.0%
      #1 of 158
    • Gemini 4 Carbon
      100.0%
      #1 of 158
    Runs and sources
    Claude Sonnet 5.5
    • max100.0%
    Gemini 4 Carbon
    • high100.0%
    Source:Epoch AI evaluation

  • SimpleQA Verified

    Best published score · Higher is better

    • Claude Sonnet 5.5
      46.5%
      #39 of 87
    • Gemini 4 Carbon
      Best77.7%
      +31.2 pts ahead · #2 of 87
    Runs and sources
    Claude Sonnet 5.5
    • max46.5%
    Gemini 4 Carbon
    • high77.7%
    Source:Epoch AI evaluation

Each model's best published result is shown; "Not reported" means no matched evaluation, not a score of zero. Arena and DeepSWE ranks are the leaderboards' own; other ranks are among catalogue models with a score. Results from different harnesses or settings are not directly comparable.

Original-source terms apply to evaluations credited to their original publishers.

Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.

Arena ratings: Arena (lmarena-ai/leaderboard-dataset), licensed CC BY 4.0. Ratings come from pairwise human votes and are only comparable within one leaderboard; the interval is Arena's 95% confidence interval. Model names are matched to the Vertical Tabs catalog.

Model specifications

Release date
September 28, 2026
October 28, 2026
Context window
1M
1M
Native context
—
—
Max output tokens
128,000
943,718
Knowledge cutoff
2026-06
2026-03
Open weights
Closed
Closed
Parameter count
—
—

API pricing · USD per 1M tokens

Input
$2.00
$2.00
Output
$10.00
$10.00
Cached input
$0.10
$0.10
Cache write
$2.50
$0.1875

Capabilities and modalities

Input modalities
text, image, file
audio, file, image, text, video
Output modalities
text
text
Tool calling
Supported
Supported
Reasoning option
Supported
Supported
Supported options
include_reasoning, max_completion_tokens, max_tokens, reasoning, reasoning_effort, response_format, stop, structured_outputs, tool_choice, tools, verbosity
include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_p
Capabilities
—
—

Architecture and availability

Architecture
—
—
Model families
—
—
Tokenizer
Claude
Gemini
Instruction format
—
—
License
—
Proprietary
Providers
5
2
Provider endpoints
20
14
Ollama variants
0
0
Metadata source
ai-research
ai-research

A dash means the field is not reported. Provider-specific performance and pricing may differ.

To give you the best experience, we use cookies for personalization, insights, and performance. Learn more.