LLM Leaderboard & AI Model Benchmarks — October 2026

As of October 28, 2026, the BenchLM AI leaderboard ranks 215 models. Gemini 4 Carbon is first by point estimate. Conditional ranges do not establish rank confidence. Open the evidence before choosing.

Search models and benchmarks

Current BenchAlign leaders

Snapshot October 28, 2026

Method

Gemini 4 CarbonEstimated
Google01
Score88.42Conditional range 74.1–100.0
Agentic
Not measured
Coding
82
Reasoning
Not measured
Multimodal
Not measured
Knowledge
88
Multilingual
Not measured
Instruction following
Not measured
Math
Not measured
Estimatedwider uncertainty
Details for Gemini 4 Carbon
Claude Opus 5.5Supported
Anthropic02
Score86.3190% interval 81.7–90.9
Agentic
90
Coding
83
Reasoning
86
Multimodal
92
Knowledge
89
Multilingual
Not measured
Instruction following
Not measured
Math
Not measured
Supportedenough independent evidence
Details for Claude Opus 5.5
GPT-6 AstraSupported
OpenAI03
Score86.0990% interval 80.6–91.6
Agentic
76
Coding
76
Reasoning
93
Multimodal
Not measured
Knowledge
86
Multilingual
Not measured
Instruction following
Not measured
Math
Not measured
Supportedenough independent evidence
Details for GPT-6 Astra

Price context: GPT-6.1 Terra retains 91% of the top score with an output price 50% lower.

Data refreshed: October 28, 2026
Supported: 77 models
Ranking method: BenchAlign v5.8

The LLM leaderboard — October 2026 rankings

Score, model size, context, price, and benchmark detail in one table. Filter state syncs to the URL so every view is shareable. Supported and Estimated labels show how much independent evidence backs each position.

Methodology Providers AI Race

CSV API/MCP

Supported means we have enough independent evidence to stand behind the row. Estimated means the rank is visible and the uncertainty is wider.

Supportedenough independent evidenceEstimatedwider uncertainty

Parameters show total billions; active counts appear when a source reports them. Size filters use total parameters: Tiny ≤4B, Small >4–<40B, Medium 40–150B, and Large >150B.

AG Agentic CO Coding RE Reasoning MM Multimodal & Grounded KN Knowledge ML Multilingual IF Instruction Following MA Math

LLM leaderboard table

#1 Gemini 4 Carbon GoogleEstimated	88.42	Not disclosed	—	$2.00 / $10.00	1527.51	Closed	—	82	—	—	88	—	—	—	100	97.48s	Proprietary	Current	Reasoning
#2 Claude Opus 5.5 AnthropicSupported	86.31	Not disclosed	1M	$4.00 / $20.00	1508.7	Closed	90	83	86	92	89	—	—	—	98	663.66s	Proprietary	Current	Reasoning
#3 GPT-6 Astra OpenAISupported	86.09	Not disclosed	1.05M	$10.00 / $50.00	1475.65	Closed	76	76	93	—	86	—	—	—	73	250.99s	Proprietary	Current	Reasoning
#4 GPT-6.1 Sol OpenAISupported	85.3	Not disclosed	1.05M	$2.00 / $10.00	1481.33	Closed	77	72	91	—	86	—	—	—	Not measured	Not measured	Proprietary	Current	Reasoning
#5 Gemini 4 Argon GoogleSupported	82.3	Not disclosed	1M	$2.00 / $10.00	1520.25	Closed	78	74	87	—	83	—	—	—	83	44.96s	Proprietary	Superseded	Reasoning
#6 Claude Sonnet 5.5 AnthropicSupported	81.86	Not disclosed	1M	$2.00 / $10.00	1477.15	Closed	69	84	82	—	82	—	—	—	138	448.12s	Proprietary	Current	Reasoning
#7 Claude Fable 5.1 AnthropicSupported	81.12	Not disclosed	1M	$10.00 / $50.00	1504.37	Closed	81	77	85	—	86	—	—	—	66	128.91s	Proprietary	Current	Reasoning
#8 Claude Fable 5 AnthropicSupported	80.82	Not disclosed	1M+	$10.00 / $50.00	1503.89	Closed	75	72	—	—	81	—	77	—	68	100.13s	Proprietary	Superseded	Reasoning
#9 GPT-6.1 Terra OpenAIEstimated	80.44	Not disclosed	1.05M	$1.00 / $5.00	—	Closed	76	—	—	92	77	—	—	—	Not measured	Not measured	Proprietary	Current	Reasoning
#10 Claude Opus 5 AnthropicSupported	79.69	Not disclosed	—	$5.00 / $25.00	1490.5	Closed	79	71	81	91	81	—	—	—	62	59.73s	Proprietary	Superseded	Reasoning
#11 GPT-6 Sol OpenAISupported	78	Not disclosed	1.05M	$2.00 / $10.00	1458.32	Closed	58	67	73	—	80	—	—	—	Not measured	Not measured	Proprietary	Superseded	Reasoning
#12 GPT-5.6 Sol OpenAISupported	74.84	Not disclosed	1.05M	$4.00 / $20.00	1484.91	Closed	69	68	76	90	76	—	89	—	94	100.66s	Proprietary	Superseded	Reasoning
#13 GPT-5.5 Pro OpenAIEstimated	72.78	Not disclosed	1M	$30.00 / $180.00	—	Closed	—	—	—	—	—	—	—	—	Not measured	Not measured	Proprietary	Established	Reasoning
#14 GPT-5.6 Terra OpenAISupported	69.83	Not disclosed	1.05M	$2.00 / $12.00	1467.34	Closed	61	62	69	85	69	—	87	—	Not measured	Not measured	Proprietary	Superseded	Reasoning
#15 Claude Opus 4.8 AnthropicSupported	69.64	Not disclosed	1M	$5.00 / $25.00	1474.44	Closed	61	61	63	88	71	—	75	84	65	35.96s	Proprietary	Superseded	Reasoning
#16 Kimi K3 Moonshot AISupported	69.38	Not disclosed	1.05M	$3.00 / $15.00	1488.19	Pending	68	59	70	90	67	—	—	—	38	56.25s	Pending	Superseded	Reasoning
#17 Muse Wonder Full MetaEstimated	68.94	Not disclosed	1M	Not listed	—	Closed	—	71	—	—	74	—	—	—	Not measured	Not measured	Proprietary	Current	Reasoning
#18 GPT-5.5 OpenAISupported	68.83	Not disclosed	1M	$5.00 / $30.00	1481.7	Closed	61	62	70	74	71	—	93	—	98	37.05s	Proprietary	Superseded	Reasoning
#19 GPT-5.4 Pro OpenAIEstimated	68.24	Not disclosed	1.05M	$30.00 / $180.00	1475.27	Closed	—	—	—	—	—	—	—	—	74	151.79s	Proprietary	Established	Reasoning
#20 Claude Opus 4.7 (Adaptive) AnthropicEstimated	68.21	Not disclosed	1M	$5.00 / $25.00	1502.21	Closed	61	56	—	51	66	—	—	—	Not measured	Not measured	Proprietary	Superseded	Reasoning
#21 Grok 4.6 xAISupported	66.66	Not disclosed	500K	$2.00 / $6.00	1455.48	Closed	70	59	61	—	68	—	—	—	61	35.43s	Proprietary	Superseded	Reasoning
#22 Gemini 3.7 Flash GoogleSupported	66.55	Not disclosed	1M	$0.75 / $3.75	1489.55	Closed	61	58	—	86	71	—	—	—	285	8.52s	Proprietary	Superseded	Reasoning
#23 Claude Haiku 5.5 AnthropicSupported	66.43	Not disclosed	1M	$0.10 / $0.50	1424.82	Closed	62	53	60	—	65	—	—	—	241	343.38s	Proprietary	Current	Reasoning
#24 GPT-5.4 OpenAISupported	66.41	Not disclosed	1.05M	$2.50 / $15.00	1465.04	Closed	54	48	64	72	68	—	91	—	89	157.51s	Proprietary	Superseded	Reasoning
#25 GPT-5.2 Pro OpenAIEstimated	66.28	Not disclosed	400K	$21.00 / $168.00	1440.15	Closed	—	—	—	—	—	—	—	—	Not measured	Not measured	Proprietary	Established	Reasoning

Showing 25 of 1002 models · 215 ranked

Source-linked · the table lists all 1002 tracked models; 215 carry a BenchAlign rank

How the score is built

What the table covers

The BenchLM LLM leaderboard 2026 ranks 215+ models and tracks 1002+ large language models side by side across 717 benchmarks — from SWE-bench and LiveCodeBench for coding to GPQA Diamond and MMLU-Pro for knowledge and reasoning. Whether you need the best AI models 2026 has to offer for agentic workflows, math, multilingual tasks, or instruction following, our AI benchmark comparison tables make it easy to see how GPT-5, Claude, Gemini, DeepSeek, Llama, and dozens of other frontier and open-source models stack up on both benchmarks and operator tradeoffs like price and context. Supported and Estimated labels separate evidence strength from score, so incomplete reporting does not automatically remove a model from the comparison.

Reading the table

Embed the live leaderboard

What changed in October 2026

Gemini 4 Carbon (Google) holds the #1 spot on the BenchAlign leaderboard at 88.42.
83 model releases tracked in October 2026.
OpenAI climbed 1 spot in the provider race.
77 Supported and 138 Estimated models appear in the public leaderboard.

Dataset journalData verified October 28, 2026

Explore benchmarks by category

Move from the overall ranking into the capability or workload that actually decides your model choice.

The evidence mapBenchmark directory

Build a head-to-head comparison

Choose any two ranked models. The comparison opens with the decision, exact units, and evidence coverage.

versus

Compare these models

Decision-ready picks

Current models that clear the ranking, freshness, and evidence thresholds for each decision.

DecisionModelCurrent measureProvenance

Best open weightHighest-scoring evidence-qualified current open-weight model.Ling 3.1 FlashInclusionAI65.47overall scoreEstimated · 11 source families
Best near-frontier value91% of the leading score at a lower output price.GPT-6.1 TerraOpenAI$5output / 1M tokensEstimated · 1 source family
Fastest measuredFastest externally measured model that also clears the ranking and evidence thresholds.Gemini 3.5 Flash-LiteGoogle461tokens / secArtificial Analysis · updated 2026-10-28
Largest useful contextLargest context among evidence-qualified current models retaining at least half of the leading score.GPT-6 AstraOpenAI1.05Mcontext windowSupported · 3 source families

Current models that clear rank, freshness and evidence thresholds

Qualification method Qualification method

Popular comparisons

Ornith-1.5-122B vs Ornith-1.5-397B Ornith-1.5-122B vs Ornith-1.5-35B-A3B ScaleDown Classify vs ScaleDown Extract ScaleDown Classify vs ScaleDown Summarize ScaleDown Classify vs ScaleDown Compress ScaleDown Extract vs ScaleDown Summarize ScaleDown Extract vs ScaleDown Compress ScaleDown Summarize vs ScaleDown Compress Gemini 4 Argon vs GPT-6 Astra

Tools and resources

AI Prompt WriterTurn the work you need into a testable prompt Prompt OptimizerGet closer to usable output with a clearer prompt LLM Pricing ComparisonCompare API costs across models Alternative FinderFind the best ChatGPT or Claude replacement Which LLM Should I Use?Get a personalized recommendation Cost CalculatorEstimate your monthly API spend Methodology & ExplainersHow benchmarks are scored and curated

How we work

A ranking is a shortlist.

Rank orders the evidence; it does not make the decision. Every row here opens to its sources so you can disagree with us.

We do not invent scores to fill a table.

A model without a public, independent row stays unranked, and says so. Vendor-run numbers are shown, labeled, and kept out of the rank.

When we were wrong, the fix is on the page.

Scoring bugs, missing coverage, and re-weightings are dated and linked, because a ranking you cannot audit is a launch post with a table.

The desk

Editorial policy

The AI race, read as a record instead of a spectacle

Track which release led its month, how provider depth changed, and whether the crown moved.

Explore the release timeline

Months: 6
Releases: 297
Lead changes: 5

Best ranked model released in Oct 2026

Gemini 4 CarbonGoogle

Score88.42

RankProviderTop-three avg.Models

01OpenAI83.937

02Anthropic83.122

03Google79.120

Six-month release record

Explore the release timeline

Scoring methodology8 weighted categories, 60 ranking benchmarks, and 373 display-only records.Inspect method

Category weights: 100%
Scored categories: 8
Ranking benchmarks: 60
Display-only: 373

Each model's overall score is a normalized weighted average of category averages. Within each category, benchmark results are normalized to a common scale and combined using weights that favor harder, less-saturated evaluations.

Confidence remains separate from score. It shows how much public benchmark evidence supports a result, while display-only benchmarks stay visible for context without receiving ranking weight.

Benchmark scores come from official provider tables, model cards, release announcements, and benchmark-native public leaderboards. External consensus signals are bounded calibration inputs and are not exposed in exported data.

Read the complete methodology

Category and weightBenchmarks that affect the scoreCoverage

Agentic22%
Terminal-Bench 3.0 · Terminal-Bench 4.0 · AA AutomationBench · AA EnterpriseOps-Gym · AA Harvey LAB v1.0 · AA Tau3 Banking · Terminal-Bench 2.0 · Terminal-Bench 2.1 · BrowseComp · HLE w/ tools · GDPval-AA · OSWorld-Verified · JobBench · MCP Atlas · Toolathlon · Toolathlon-Verified · AutomationBench · Agents' Last Exam · DeepSearchQA · BFCL v4 · τ³-bench results · Terminal-Bench 2.1 (Vals)
22 weighted · 103 context-only
Coding20%
SWE-Rebench · SWE-bench Pro · VulcanBench v3 · FrontierCode 1.1 Main · SWE Multilingual · FrontierSWE v2 · SciCode · AA-SciCode · LiveCodeBench (Vals) · DeepSWE
10 weighted · 57 context-only
Reasoning17%
LongBench v2 · MRCRv2 · ARC-AGI-2 · ARC-AGI-3 · AA-LCR · CritPt
6 weighted · 31 context-only
Multimodal12%
MMMU-Pro · AA-MMMU-Pro · OfficeQA Pro · CharXiv
4 weighted · 71 context-only
Knowledge12%
GPQA · SuperGPQA · MMLU-Pro · HLE · AA-GPQA Diamond · AA-HLE · AA-Omniscience Accuracy · HLE w/o tools · GPQA Diamond (Vals) · MMLU-Pro (Vals)
10 weighted · 54 context-only
Multilingual7%
MMLU-ProX
1 weighted · 13 context-only
Instruction Following5%
IFBench · AA-IFBench
2 weighted · 15 context-only
Math5%
AIME26 · HMMT Feb 2026 · FrontierMath v2 (Tiers 1-3) · FrontierMath v2 (Tier 4) · USAMO 2026
5 weighted · 29 context-only

Scoring methodology · BenchAlign v5.8

Inspect method