LLM leaderboard, August 2026

LLM Leaderboard & AI Model Benchmarks — August 2026

Compare frontier AI models by quality, cost, and context. 104 Supported and 113 Estimated models among 388 tracked LLMs — 430 benchmarks, real pricing, and runtime data in one place.

Data refreshed: August 13, 2026
Supported: 104 models
Ranking method: BenchAlign v5.2

Radar · latest confirmed releases

3 source-linked updates

Free Radar Brief ranks up to five confirmed market changes and keeps the original source attached. No card required.

Browse public updates Start the free Radar Brief

Latest

DeepSeek V4 Pro 0813 released

DeepSeek · Aug 13, 2026

Source confirmed

Gemini 3.7 Flash released

Google · Aug 13, 2026

Grok 4.6 released

xAI · Aug 12, 2026

Decision-ready picks

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

Qualification method

DecisionModelCurrent measureProvenance

Best open weightHighest-scoring evidence-qualified current open-weight model.Qwen3.8 MaxAlibaba79.63overall scoreSupported · 2 source families
Best near-frontier value97% of the leading score at a lower output price.Kimi K3Moonshot AI$15output / 1M tokensSupported · 2 source families
Fastest measuredFastest externally measured model that also clears the ranking and evidence thresholds.Ling 3.0 FlashInclusionAI394tokens / secArtificial Analysis · updated 2026-08-12
Largest useful contextLargest context among evidence-qualified current models retaining at least half of the leading score.Grok 4.20xAI2Mcontext windowEstimated · 1 source family

Qualification method

The LLM leaderboard — August 2026 rankings

Benchmarks, pricing, runtime signals, and context window 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 JSON


1 Claude Mythos 5 AnthropicSupported	Anthropic	Closed	Current	Reasoning	1M+	$10.00 / $50.00	Not measured	Not measured	82.87	75	80	—	86	94	—	—	—	—
2 Claude Opus 5 AnthropicSupported	Anthropic	Closed	Current	Reasoning	—	$5.00 / $25.00	54	52.35s	82.8	80	78	91	88	93	—	—	—	1493.6
3 Claude Fable 5 AnthropicSupported	Anthropic	Closed	Current	Reasoning	1M+	$10.00 / $50.00	63	99.62s	82.62	75	80	—	63	67	—	—	—	1505.66
4 GPT-5.6 Sol OpenAISupported	OpenAI	Closed	Current	Reasoning	1.05M	$5.00 / $30.00	62	145.22s	81.79	68	79	93	85	84	—	—	97	1480.8
5 Kimi K3 Moonshot AISupported	Moonshot AI	Closed	Current	Reasoning	1.05M	$3.00 / $15.00	41	51.31s	80.21	74	78	—	88	85	—	—	—	1485.33
6 Qwen3.8 Max AlibabaSupported Self-host	Alibaba	Open	Current	Reasoning	1M	Not listed	48	44.73s	79.63	78	66	96	88	67	—	98	—	1490.7
7 Claude Opus 4.8 AnthropicSupported	Anthropic	Closed	Superseded	Reasoning	1M	$5.00 / $25.00	56	18.31s	76.84	60	72	76	88	88	—	—	67	1482.43
8 Muse Spark 1.1 MetaSupported	Meta	Closed	Superseded	Reasoning	1M	Not listed	217	12.12s	76.57	65	64	—	76	93	—	—	—	1488.67
9 Gemini 3.6 Flash GoogleSupported	Google	Closed	Superseded	Reasoning	1M	$1.50 / $7.50	238	17.93s	75.25	46	65	—	73	68	—	—	—	1483.64
10 Grok 4.5 xAISupported	xAI	Closed	Superseded	Reasoning	500K	$2.00 / $6.00	56	8.68s	75.1	61	57	60	—	72	—	—	—	1469.43
11 GPT-5.5 OpenAIEstimated	OpenAI	Closed	Superseded	Reasoning	1M	$5.00 / $30.00	69	88.09s	73.09	64	71	87	68	79	—	—	71	1481.36
12 GPT-5.4 OpenAISupported	OpenAI	Closed	Superseded	Reasoning	1.05M	$2.50 / $15.00	123	134.85s	73.05	57	58	78	67	77	—	—	66	1465.51
13 GPT-5.6 Terra OpenAIEstimated	OpenAI	Closed	Current	Reasoning	1.05M	$2.00 / $12.00	Not measured	Not measured	72.55	62	66	86	75	82	—	—	97	—
14 Claude Opus 4.7 (Adaptive) AnthropicEstimated	Anthropic	Closed	Superseded	Reasoning	1M	$5.00 / $25.00	Not measured	Not measured	72.15	64	65	79	48	83	—	—	—	1502.03
15 Claude Opus 4.7 AnthropicSupported	Anthropic	Closed	Current	Standard	1M	$5.00 / $25.00	47	15.93s	71.57	56	68	—	—	66	—	—	62	1493.75
16 Qwen3.7 Max AlibabaSupported	Alibaba	Closed	Superseded	Reasoning	1M	Not listed	205	14.13s	71.31	39	54	87	—	72	100	93	82	1473.9
17 Muse Spark MetaSupported	Meta	Closed	Superseded	Reasoning	262K	Not listed	Not measured	Not measured	70.54	65	63	52	76	75	—	—	56	1487.95
18 MiMo-V2.5-Pro XiaomiSupported	Xiaomi	Closed	Current	Reasoning	1M	Not listed	Not measured	Not measured	68.91	41	55	—	—	71	—	—	—	1467.5
19 MiniMax M3 MiniMaxSupported Self-host	MiniMax	Open	Current	Standard	1M	$0.30 / $1.20	80	26.71s	68.37	40	48	—	47	64	—	—	—	1443.25
20 Claude Opus 4.6 AnthropicSupported	Anthropic	Closed	Superseded	Standard	1M	$5.00 / $25.00	38	2.52s	67.73	57	61	—	62	83	—	—	60	1497.59
21 Hy3 TencentSupported Self-host	Tencent	Open	Current	Reasoning	256K	$0.00 / $0.00	64	34.32s	67.71	61	63	—	—	—	—	—	—	1456.32
22 Gemini 3 Pro GoogleSupported	Google	Closed	Established	Standard	2M	$2.00 / $12.00	109	32.65s	67.15	58	62	42	73	73	—	—	56	1485.45
23 GPT-5.6 Luna OpenAIEstimated	OpenAI	Closed	Current	Reasoning	1.05M	$0.20 / $1.20	Not measured	Not measured	66.92	54	72	66	66	82	—	—	97	—
24 Inkling Thinking Machines LabSupported Self-host	Thinking Machines Lab	Open	Current	Standard	1M	$1.87 / $4.68	76	28.30s	66.78	48	43	—	51	69	—	91	79	1442.6
25 MiMo-V2-Pro XiaomiSupported	Xiaomi	Closed	Superseded	Reasoning	1M	Not listed	Not measured	Not measured	66.64	—	61	—	—	—	—	—	—	1448.32

Showing 25 of 382

Reading the table

Use rank to shortlist, then inspect the evidence

Embed the live leaderboard

The BenchLM LLM leaderboard 2026 ranks 217+ models and tracks 388+ large language models side by side across 430 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.

Dataset journal

What changed in August 2026

Claude Mythos 5 (Anthropic) holds the #1 spot on the BenchAlign leaderboard at 82.87.
14 model releases tracked in August 2026.
Meta climbed 2 spots in the provider race.
104 Supported and 113 Estimated models appear in the public leaderboard.

Data verified August 13, 2026

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

The evidence map

Explore benchmarks by category

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

Agentic82Tools, browsers, and long-running work Coding63Repository tasks and code generation Multimodal65Documents, images, and grounded work Reasoning27Long-context and abstract reasoning Knowledge46Professional and scientific knowledge Instruction following4Constraint and format reliability Multilingual13Cross-language accuracy and coverage Math32Contest and frontier mathematics

Six-month release record

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.

Months: 6
Releases: 159
Lead changes: 5

Explore the release timeline

Best ranked model released in Aug 2026

Qwen3.8 MaxAlibaba

Score79.63

RankProviderTop-three avg.Models

01Anthropic82.820

02OpenAI7638

03Alibaba72.221

Popular comparisons

Claude Opus 5 vs GPT-5.6 Sol GPT-5.6 Sol vs Claude Sonnet 5 Kimi K3 vs GPT-5.6 Sol GPT-5.6 Sol vs Claude Opus 4.8 GPT-5.6 Terra vs Gemini 3.5 Flash GPT-5.6 Luna vs Gemini 3.5 Flash Gemini 3.6 Flash vs Gemini 3.5 Flash Gemini 3.5 Flash-Lite vs Gemini 3.1 Flash-Lite Claude Sonnet 5 vs Claude Opus 4.8

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 BenchLM scores and curates benchmarks

Scoring methodology8 weighted categories, 27 ranking benchmarks, and 305 display-only records.Inspect method

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.

Data comes from OpenBench, official model papers, and public leaderboards. External consensus signals are bounded calibration inputs and are not exposed in exported data.

Read the complete methodology

Category weights: 100%
Scored categories: 8
Ranking benchmarks: 27
Display-only: 305

Category and weightBenchmarks that affect the scoreCoverage

Agentic22%
Terminal-Bench 2.0 · BrowseComp · OSWorld-Verified
3 weighted · 79 context-only
Coding20%
SWE-bench Verified · SWE-Rebench · LiveCodeBench · SWE-bench Pro · SciCode
5 weighted · 58 context-only
Reasoning17%
LongBench v2 · MRCRv2 · ARC-AGI-2
3 weighted · 24 context-only
Multimodal12%
MMMU-Pro · OfficeQA Pro · CharXiv
3 weighted · 62 context-only
Knowledge12%
GPQA · SuperGPQA · MMLU-Pro · HLE · SimpleQA
5 weighted · 41 context-only
Multilingual7%
MMLU-ProX
1 weighted · 12 context-only
Instruction Following5%
IFEval · IFBench
2 weighted · 2 context-only
Math5%
AIME26 · HMMT Feb 2026 · FrontierMath v2 (Tiers 1-3) · FrontierMath v2 (Tier 4) · USAMO 2026
5 weighted · 27 context-only

Browse BenchLM

Tools & calculators

Tools