OutYet.ai

Benchmarks

Compare AI models

Pick two models to see release dates, benchmark scores, and community ratings side by side.

Frontier benchmarks

Frontier board

6 releases. Five ways to compare.

Selected for recent releases and benchmark coverage, not a single overall rank. Each column sorts the same models by its own score. No reading means no bar.

○ Independent evaluation · ◆ Vendor or self-reported, labeled per reading.

Coding

BenchLM · 0 to 100

  1. Claude Fable 5.176.82Independent

    Anthropic

  2. GPT-6 Astra75.43Independent

    OpenAI

  3. Gemini 4 Argon71.22Independent

    Google DeepMind

  4. GPT-6.1 Sol70.86Independent

    OpenAI

  5. Claude Opus 570.68Independent

    Anthropic

  6. Kimi K358.99Independent

    Moonshot AI

Agentic

BenchLM · 0 to 100

  1. Claude Fable 5.180.86Independent

    Anthropic

  2. Claude Opus 579.14Independent

    Anthropic

  3. GPT-6.1 Sol75.88Independent

    OpenAI

  4. Gemini 4 Argon75.13Independent

    Google DeepMind

  5. GPT-6 Astra72.21Independent

    OpenAI

  6. Kimi K370.49Independent

    Moonshot AI

Reasoning

BenchLM · 0 to 100

  1. GPT-6 Astra93.3Independent

    OpenAI

  2. GPT-6.1 Sol90.5Independent

    OpenAI

  3. Claude Fable 5.185.2Independent

    Anthropic

  4. Claude Opus 581Independent

    Anthropic

  5. Kimi K370Independent

    Moonshot AI

  6. Gemini 4 ArgonNot reported

    Google DeepMind

Knowledge

BenchLM · 0 to 100

  1. Claude Fable 5.186.38Independent

    Anthropic

  2. GPT-6 Astra85.67Independent

    OpenAI

  3. GPT-6.1 Sol85.48Independent

    OpenAI

  4. Claude Opus 581.62Independent

    Anthropic

  5. Gemini 4 Argon78.73Independent

    Google DeepMind

  6. Kimi K368.5Independent

    Moonshot AI

Arena

LM Arena · Elo

  1. Gemini 4 Argon1,524.77Independent

    Google DeepMind

  2. Claude Fable 5.11,500.57Independent

    Anthropic

  3. Claude Opus 51,489.85Independent

    Anthropic

  4. Kimi K31,488.1Independent

    Moonshot AI

  5. GPT-6.1 Sol1,483.96Independent

    OpenAI

  6. GPT-6 Astra1,476.72Independent

    OpenAI

Score links open the source; hover for capture details. Bar lengths start at zero. Arena lengths use the full public board maximum.

The shape of the scores

First three featured models. Toggle a name to isolate a line. Missing readings leave gaps, not zeroes.

CodingAgenticReasoningKnowledgeArena050100Claude Fable 5.1, Coding: 76.82. Independent. Captured 2026-10-14.Claude Fable 5.1, Agentic: 80.86. Independent. Captured 2026-10-14.Claude Fable 5.1, Reasoning: 85.2. Independent. Captured 2026-10-14.Claude Fable 5.1, Knowledge: 86.38. Independent. Captured 2026-10-14.Claude Fable 5.1, Arena: 1,500.57. Independent. Captured 2026-10-14.GPT-6 Astra, Coding: 75.43. Independent. Captured 2026-10-14.GPT-6 Astra, Agentic: 72.21. Independent. Captured 2026-10-14.GPT-6 Astra, Reasoning: 93.3. Independent. Captured 2026-10-14.GPT-6 Astra, Knowledge: 85.67. Independent. Captured 2026-10-14.GPT-6 Astra, Arena: 1,476.72. Independent. Captured 2026-10-14.GPT-6.1 Sol, Coding: 70.86. Independent. Captured 2026-10-14.GPT-6.1 Sol, Agentic: 75.88. Independent. Captured 2026-10-14.GPT-6.1 Sol, Reasoning: 90.5. Independent. Captured 2026-10-14.GPT-6.1 Sol, Knowledge: 85.48. Independent. Captured 2026-10-14.GPT-6.1 Sol, Arena: 1,483.96. Independent. Captured 2026-10-14.

3 of 3 series shown.

Axes share drawing space, not meaning. Arena is Elo divided by the current board maximum; it is not a percentage accuracy.

The newest arrival

Gemini 4 Argon

Released 2026-10-14 · At least two readings

No predecessor is linked in the catalog yet. We do not infer a generation jump from a name.

Latest release with linked lineage

Claude Fable 5.1 vs Claude Fable 5

Coding
+5.56pts
71.26 to 76.82 Independent2026-10-14 / 2026-10-14
Agentic
+4.41pts
76.45 to 80.86 Independent2026-10-14 / 2026-10-14
Knowledge
+4.29pts
82.09 to 86.38 Independent2026-10-14 / 2026-10-14
Arena
-2.81Elo
1,503.38 to 1,500.57 Independent2026-10-14 / 2026-10-14

Latest snapshots, not a controlled before-and-after trial. Capture dates and evaluation setups may differ.

View as table
Frontier board: original scores. Not reported is missing, not zero.
ModelMetricScore / provenanceCapturedSource label / votes
Claude Fable 5.1CodingBenchLM.ai76.82Independent2026-10-14Claude Fable 5.1
GPT-6 AstraCodingBenchLM.ai75.43Independent2026-10-14GPT-6 Astra
Gemini 4 ArgonCodingBenchLM.ai71.22Independent2026-10-14Gemini 4 Argon
GPT-6.1 SolCodingBenchLM.ai70.86Independent2026-10-14GPT-6.1 Sol
Claude Opus 5CodingBenchLM.ai70.68Independent2026-10-14Claude Opus 5
Kimi K3CodingBenchLM.ai58.99Independent2026-10-14Kimi K3
Claude Fable 5.1AgenticBenchLM.ai80.86Independent2026-10-14Claude Fable 5.1
Claude Opus 5AgenticBenchLM.ai79.14Independent2026-10-14Claude Opus 5
GPT-6.1 SolAgenticBenchLM.ai75.88Independent2026-10-14GPT-6.1 Sol
Gemini 4 ArgonAgenticBenchLM.ai75.13Independent2026-10-14Gemini 4 Argon
GPT-6 AstraAgenticBenchLM.ai72.21Independent2026-10-14GPT-6 Astra
Kimi K3AgenticBenchLM.ai70.49Independent2026-10-14Kimi K3
GPT-6 AstraReasoningBenchLM.ai93.3Independent2026-10-14GPT-6 Astra
GPT-6.1 SolReasoningBenchLM.ai90.5Independent2026-10-14GPT-6.1 Sol
Claude Fable 5.1ReasoningBenchLM.ai85.2Independent2026-10-14Claude Fable 5.1
Claude Opus 5ReasoningBenchLM.ai81Independent2026-10-14Claude Opus 5
Kimi K3ReasoningBenchLM.ai70Independent2026-10-14Kimi K3
Gemini 4 ArgonReasoningBenchLM.aiNot reportedNot reportedNot reported
Claude Fable 5.1KnowledgeBenchLM.ai86.38Independent2026-10-14Claude Fable 5.1
GPT-6 AstraKnowledgeBenchLM.ai85.67Independent2026-10-14GPT-6 Astra
GPT-6.1 SolKnowledgeBenchLM.ai85.48Independent2026-10-14GPT-6.1 Sol
Claude Opus 5KnowledgeBenchLM.ai81.62Independent2026-10-14Claude Opus 5
Gemini 4 ArgonKnowledgeBenchLM.ai78.73Independent2026-10-14Gemini 4 Argon
Kimi K3KnowledgeBenchLM.ai68.5Independent2026-10-14Kimi K3
Gemini 4 ArgonArenaLM Arena1,524.77Independent2026-10-14gemini-4-argon-high4,903 votes
Claude Fable 5.1ArenaLM Arena1,500.57Independent2026-10-14claude-fable-5.1-max14,121 votes
Claude Opus 5ArenaLM Arena1,489.85Independent2026-10-14claude-opus-5-high65,663 votes
Kimi K3ArenaLM Arena1,488.1Independent2026-10-14kimi-k3-max32,036 votes
GPT-6.1 SolArenaLM Arena1,483.96Independent2026-10-14gpt-6.1-sol-max5,762 votes
GPT-6 AstraArenaLM Arena1,476.72Independent2026-10-14gpt-6-astra-max11,674 votes
Additional readings for these models. Separate benchmarks, not substitutes for the five chart axes.
ModelMetric / sourceScore / provenanceCaptured
Claude Fable 5.1Humanity's Last ExamEpoch AI46.5Independent2026-10-14
Claude Fable 5.1GDP.pdfEpoch AI29.6Independent2026-10-14
GPT-6 AstraHumanity's Last ExamEpoch AI54.8Independent2026-10-14
Claude Opus 5MultimodalGroundedBenchLM.ai92.2Independent2026-10-14
Claude Opus 5GDP.pdfEpoch AI24Independent2026-10-14
Kimi K3MultimodalGroundedBenchLM.ai92.6Independent2026-10-14
Open page →
Provider
OpenAI
-
Out yet?
Yes
-
Status
Available
-
Release date
Jul 9, 2026
-
LM Arena?LM Arena's Elo-style rating from blind head-to-head votes: people compare two anonymous model answers and pick the better one. Higher is better.
1452.5Verify at LM ArenaOct 14, 2026
-
Agentic?BenchLM.ai's score for multi-step agentic work - planning, tool use, and acting autonomously, normalized 0–100 across multiple benchmarks. Higher is better.
56.26Verify at BenchLM.aiOct 14, 2026
-
Coding?BenchLM.ai's score for code generation and software-engineering tasks, normalized 0–100 across multiple benchmarks. Higher is better.
61.67Verify at BenchLM.aiOct 14, 2026
-
Knowledge?BenchLM.ai's score for factual knowledge and question answering, normalized 0–100 across multiple benchmarks. Higher is better.
64.53Verify at BenchLM.aiOct 14, 2026
-
MultimodalGrounded?BenchLM.ai's score for understanding grounded in images and documents, normalized 0–100 across multiple benchmarks. Higher is better.
79.6Verify at BenchLM.aiOct 14, 2026
-
Reasoning?BenchLM.ai's score for logic and multi-step reasoning problems, normalized 0–100 across multiple benchmarks. Higher is better.
58.2Verify at BenchLM.aiOct 14, 2026
-
GDP.pdf?Epoch AI's GDP.pdf benchmark: extracting and analyzing information from real-world PDF documents. Higher is better.Epoch AI
22.7Verify at Epoch AIAug 25, 2026
-
HealthBench?OpenAI's HealthBench evaluation of healthcare conversation quality. This value is vendor-reported; higher is better.openai-system-card
55.8Verify at openai-system-cardJul 10, 2026
-
HealthBench Consensus?OpenAI's HealthBench Consensus evaluation. This value is vendor-reported; higher is better.openai-system-card
95.1Verify at openai-system-cardJul 10, 2026
-
HealthBench Hard?The harder subset of OpenAI's HealthBench healthcare evaluation. This value is vendor-reported; higher is better.openai-system-card
32Verify at openai-system-cardJul 10, 2026
-
HealthBench Professional?OpenAI's HealthBench Professional evaluation of model responses to realistic healthcare conversations. This value is vendor-reported; higher is better.openai-system-card
55.7Verify at openai-system-cardJul 10, 2026
-
Vibe rating?OutYet's community rating: signed-in users score the model 1–10. Shown as the average and the number of votes.
-
-
Predecessor
-
-
Successor
-
-

Benchmark scores are mirrored from third-party sources and captured on the dates shown. Numbers from different benchmarks, sources, or capture dates are not directly comparable.

Data from BenchLM.ai · Epoch AI, “AI Benchmarking Hub” (CC BY 4.0).