Skip to main content

Compare LLMs without mixing incompatible test setups

Compare 50 leading language models across 1,225 head-to-head matchups. Every result includes shared benchmark cohorts, current API pricing, key specifications, and source links.

models
50
head-to-head matchups
1,225
active benchmarks
89
active results
14,680

Cross-benchmark view

Overall LLM leaderboard

The index uses 10 benchmark cohorts with comparable documented test setups. It converts the scores into rank percentiles and averages them with equal weight over the cohorts a model appears in. The 86 models measured in at least 4 of them receive an overall rank, and each row states how many carry it. Ranks based on fewer than half of the cohorts are marked as provisional, because they usually move once more results are published. Models below that threshold are omitted, and missing data is not treated as a penalty.

1.Claude Fable 5.193.5 / 100, Rank 110 of 10 cohorts
2.Claude Fable 593.2 / 10010 of 10 cohorts
3.Claude Opus 591.3 / 10010 of 10 cohorts
4.Claude Sonnet 5.587.9 / 1004 of 10 cohortsProvisional
5.Claude Opus 5.585.1 / 1004 of 10 cohortsProvisional
6.Kimi K385.1 / 10010 of 10 cohorts
7.Gemini 3.7 Flash83.1 / 1009 of 10 cohorts
8.GPT-5.6 Sol81.9 / 10010 of 10 cohorts
9.Claude Opus 4.878.8 / 10010 of 10 cohorts
10.Gemini 3.8 Flash77.4 / 10010 of 10 cohorts
11.Qwen 3.8 Max 090276.5 / 10010 of 10 cohorts
12.GPT-5.575.8 / 10010 of 10 cohorts
13.Claude Opus 4.772.6 / 1009 of 10 cohorts
14.Gemini 3.6 Flash70.4 / 10010 of 10 cohorts
15.Grok 4.669.7 / 1009 of 10 cohorts
16.Qwen 3.7 Max68.7 / 1007 of 10 cohorts
17.GPT-5.268.3 / 1008 of 10 cohorts
18.Gemini 3.1 Pro Preview67.5 / 10010 of 10 cohorts
19.GPT-5.6 Terra67.5 / 10010 of 10 cohorts
20.Gemini 3.5 Flash67.4 / 10010 of 10 cohorts
Show all 86 ranked models
21.Gemini 3 Pro66.3 / 1008 of 10 cohorts
22.GLM-5.365.3 / 1008 of 10 cohorts
23.GPT-5.6 Luna64.4 / 1009 of 10 cohorts
24.Grok 4.763.7 / 1005 of 10 cohorts
25.DeepSeek-V4.1-Flash62.8 / 1005 of 10 cohorts
26.Gemini 3 Flash Preview62.7 / 1009 of 10 cohorts
27.GPT-5.462.7 / 1008 of 10 cohorts
28.GLM-5.3-Flash62.4 / 1008 of 10 cohorts
29.Claude Sonnet 561.6 / 10010 of 10 cohorts
30.Kimi K2.659.2 / 10010 of 10 cohorts
31.Grok 4.558.6 / 1009 of 10 cohorts
32.MiniMax M358.3 / 10010 of 10 cohorts
33.GPT-557.6 / 1008 of 10 cohorts
34.Claude Sonnet 4.655.9 / 1009 of 10 cohorts
35.Qwen 3.8 27B53.7 / 1009 of 10 cohorts
36.GLM-5.252.6 / 1007 of 10 cohorts
37.GPT-6 Luna50.7 / 1004 of 10 cohortsProvisional
38.Nemotron 3 Ultra49.8 / 1006 of 10 cohorts
39.GPT-6 Sol49.6 / 1004 of 10 cohortsProvisional
40.DeepSeek-V4-Pro49.6 / 1008 of 10 cohorts
41.Inkling Small48.5 / 1009 of 10 cohorts
42.Kimi K2.548.2 / 1008 of 10 cohorts
43.Inkling47.8 / 10010 of 10 cohorts
44.o346.9 / 1008 of 10 cohorts
45.GPT-5 mini46.1 / 1008 of 10 cohorts
46.GLM-546.0 / 1005 of 10 cohorts
47.GLM-5.144.2 / 1008 of 10 cohorts
48.Claude Opus 4.543.7 / 1008 of 10 cohorts
49.GPT-5.4 mini42.3 / 1007 of 10 cohorts
50.MiMo-V2.5-Pro42.2 / 1007 of 10 cohorts
51.Grok 4.20 Reasoning39.3 / 1009 of 10 cohorts
52.Grok 338.7 / 1005 of 10 cohorts
53.Gemini 3.1 Flash-Lite38.5 / 1009 of 10 cohorts
54.Gemini 3.5 Flash-Lite38.2 / 10010 of 10 cohorts
55.Grok 4.336.8 / 10010 of 10 cohorts
56.MiniMax M2.135.5 / 1006 of 10 cohorts
57.MiniMax M2.735.0 / 1007 of 10 cohorts
58.MiMo-V2.533.6 / 1009 of 10 cohorts
59.GLM-4.733.4 / 1006 of 10 cohorts
60.o4-mini33.3 / 1008 of 10 cohorts
61.GPT-4.132.6 / 1006 of 10 cohorts
62.Ling 3.0 Flash32.2 / 1008 of 10 cohorts
63.o131.4 / 1006 of 10 cohorts
64.GPT OSS 120B30.6 / 1005 of 10 cohorts
65.Kimi K2 Thinking28.8 / 1005 of 10 cohorts
66.MiniMax M2.528.2 / 1005 of 10 cohorts
67.DeepSeek-R127.3 / 1004 of 10 cohortsProvisional
68.GLM-4.625.8 / 1005 of 10 cohorts
69.GLM-4.525.6 / 1005 of 10 cohorts
70.Kimi K2 071124.4 / 1005 of 10 cohorts
71.GPT-5.4 nano24.0 / 1009 of 10 cohorts
72.DeepSeek-V3.222.4 / 1004 of 10 cohortsProvisional
73.Mistral Large 319.8 / 1007 of 10 cohorts
74.GPT-4.1 mini19.7 / 1006 of 10 cohorts
75.o3-mini19.7 / 1005 of 10 cohorts
76.DeepSeek-V3 032418.8 / 1005 of 10 cohorts
77.DeepSeek-V317.8 / 1004 of 10 cohortsProvisional
78.GPT OSS 20B15.1 / 1005 of 10 cohorts
79.GPT-5 nano12.4 / 1008 of 10 cohorts
80.Command A10.5 / 1005 of 10 cohorts
81.Mistral Medium 3.59.5 / 1006 of 10 cohorts
82.Llama 4 Scout6.1 / 1008 of 10 cohorts
83.Mistral Small 3.14.7 / 1006 of 10 cohorts
84.GPT-4.1 nano4.0 / 1006 of 10 cohorts
85.Command R+2.5 / 1004 of 10 cohortsProvisional
86.GPT-4o mini2.3 / 1005 of 10 cohorts
Vals AILegalBenchVals AIGPQA DiamondVals AILiveCodeBench v6Vals AIMMLU-ProVals AITaxEval v2Vals AIMedScribeVals AISAGEVals AITerminal-Bench 2.1Vals AIMMMU-ProVals AIPublic Benefits Bench v1.1

This is a transparent orientation index, not an additional benchmark. It does not include price, speed, context length, or results outside the listed 10 benchmark cohorts.

Source entries with incompatible documented configurations are excluded from the composite instead of being reduced to their strongest result.

Current leaders

Which LLMs lead the most important benchmarks?

These five charts keep the benchmarks separate. Each ranking uses one published leaderboard cohort, one task, and one metric. Models without an unambiguous published result in that cohort are omitted instead of receiving an estimated score.

Overall preference

LMArena Text, Style Control

Style-controlled rating from anonymous pairwise comparisons.

1.Kimi K31,488 points, Rank 1
2.Gemini 3.1 Pro Preview1,486.7 points
3.GPT-5.6 Sol1,483.2 points
4.Gemini 3.6 Flash1,481.9 points
5.GLM-5.21,475.6 points
6.MiMo-V2.5-Pro1,467.4 points
7.Grok 4.51,465.4 points
8.GLM-5.11,465.1 points
9.GPT-5.6 Terra1,465 points
10.Claude Sonnet 51,461.5 points
11.Kimi K2.61,460.9 points
12.Gemini 3.5 Flash-Lite1,456 points
13.GPT-5.6 Luna1,454 points
14.Grok 4.61,453.4 points
15.GPT-5.4 mini1,447 points
16.Grok 4.31,441.6 points
17.MiMo-V2.51,433.5 points
18.Mistral Medium 3.51,426.1 points
19.Mistral Large 31,413.3 points
LMArenaUpdated Sep 25, 2026 · 19 of 50 comparison models with an unambiguous result

The published leaderboard documents model-specific test settings. 6 comparison models have more than one incompatible setup. They are omitted from this ranking instead of selecting the best result.

Reasoning

GPQA Diamond

Difficult scientific questions from the GPQA Diamond dataset.

1.Gemini 3.1 Pro Preview95.5 %, Rank 1
2.GPT-5.6 Sol95.2 %
3.Grok 4.694.7 %
4.Gemini 3.8 Flash94.4 %
5.Gemini 3.7 Flash93.9 %
6.Qwen 3.8 Max 090293.7 %
7.Claude Fable 5.193.4 %
7.Claude Opus 593.4 %
7.Gemini 3.6 Flash93.4 %
10.Claude Fable 593.2 %
10.GPT-5.593.2 %
12.Grok 4.592.9 %
12.Kimi K392.9 %
14.Gemini 3.5 Flash92.7 %
15.Claude Opus 4.892.4 %
16.GPT-5.491.7 %
16.GPT-5.6 Luna91.7 %
18.Grok 4.391.4 %
19.GPT-5.6 Terra90.9 %
20.Claude Opus 4.790.2 %
21.DeepSeek-V4-Pro89.4 %
22.Kimi K2.689.1 %
23.Claude Sonnet 588.9 %
24.GLM-5.388.1 %
25.GLM-5.3-Flash86.4 %
26.Claude Sonnet 4.685.6 %
26.GLM-5.285.6 %
28.GLM-5.184.5 %
29.Gemini 3.5 Flash-Lite83.8 %
30.GPT-5.4 mini83.1 %
31.MiMo-V2.5-Pro82.6 %
32.MiMo-V2.581.6 %
33.Claude Opus 4.579.5 %
34.Mistral Large 368.4 %
35.Mistral Medium 3.534.8 %
Vals AIUpdated Sep 1, 2026 · 35 of 50 comparison models with an unambiguous result

Coding

Vibe Code Bench v1.1

Practical programming tasks from Vibe Code Bench v1.1.

1.Claude Sonnet 5.592.4 %, Rank 1
2.Claude Fable 590.4 %
3.Claude Opus 5.590.3 %
4.Claude Fable 5.190.3 %
5.GPT-6 Astra89.6 %
6.Claude Opus 588.4 %
7.GPT-6 Sol87.8 %
8.Grok 4.786.2 %
9.Kimi K385 %
10.DeepSeek-V4.1-Flash84.7 %
11.Claude Opus 4.882.7 %
12.GPT-6 Luna81.6 %
13.Claude Sonnet 581.3 %
14.GPT-5.6 Sol80.5 %
15.Gemini 3.8 Flash78.7 %
16.GLM-5.378.1 %
17.GPT-5.6 Luna77.1 %
18.Grok 4.676.2 %
19.GPT-5.6 Terra74.6 %
20.Claude Opus 4.771 %
21.Gemini 3.7 Flash70.4 %
22.GPT-5.569.8 %
23.Grok 4.569 %
24.Qwen 3.8 Max 090264.7 %
25.Gemini 3.6 Flash64 %
26.GLM-5.264 %
27.Claude Opus 4.657.6 %
28.Claude Sonnet 4.651.5 %
29.DeepSeek-V4-Pro49.9 %
30.Gemini 3.5 Flash48.7 %
31.GPT-5.4 mini48 %
32.MiMo-V2.542.2 %
33.Kimi K2.637.9 %
34.Gemini 3.5 Flash-Lite37.2 %
35.MiMo-V2.5-Pro34.1 %
36.Gemini 3.1 Pro Preview32 %
37.GLM-5.131.5 %
38.GLM-5.3-Flash30.8 %
39.Grok 4.319.4 %
40.Mistral Medium 3.52.9 %
Vals AIUpdated Sep 27, 2026 · 40 of 50 comparison models with an unambiguous result

The published leaderboard documents model-specific test settings. 1 comparison model has more than one incompatible setup. It is omitted from this ranking instead of selecting the best result.

Agents

Terminal-Bench 2.1

Multi-step tasks completed in a real terminal environment.

1.Claude Opus 5.587.6 %, Rank 1
2.GPT-6 Astra87.3 %
3.GPT-5.6 Sol85.8 %
4.Claude Fable 5.185 %
5.Claude Opus 584.6 %
6.Claude Sonnet 5.583.1 %
6.GPT-6 Sol83.1 %
8.Gemini 3.8 Flash81.3 %
9.Kimi K380.9 %
10.Claude Fable 580.5 %
11.GPT-5.6 Luna79 %
12.Grok 4.678.3 %
13.Gemini 3.7 Flash77.5 %
13.GPT-5.6 Terra77.5 %
15.GPT-5.576.4 %
16.Claude Sonnet 574.5 %
16.DeepSeek-V4.1-Flash74.5 %
18.Gemini 3.5 Flash74.2 %
19.Gemini 3.6 Flash73.8 %
20.Grok 4.773.4 %
21.GPT-6 Luna73 %
22.Claude Opus 4.871.9 %
23.GLM-5.371.5 %
24.Gemini 3.1 Pro Preview70.8 %
25.Claude Opus 4.768.5 %
26.GLM-5.267.8 %
26.Grok 4.567.8 %
28.Qwen 3.8 Max 090267.4 %
29.GLM-5.3-Flash62.9 %
30.MiMo-V2.560.7 %
31.Claude Sonnet 4.657.3 %
31.MiMo-V2.5-Pro57.3 %
33.GLM-5.156.9 %
34.GPT-5.4 mini54.7 %
35.Kimi K2.653.6 %
36.DeepSeek-V4-Pro50.2 %
36.Gemini 3.5 Flash-Lite50.2 %
38.Grok 4.341.9 %
39.Mistral Medium 3.539 %
Vals AIUpdated Sep 27, 2026 · 39 of 50 comparison models with an unambiguous result

Finance

Finance Agent (v2)

Research and analysis tasks from Finance Agent v2.

1.Gemini 3.8 Flash61.4 %, Rank 1
2.Gemini 3.7 Flash59 %
3.Claude Fable 5.158.9 %
4.Claude Opus 558.6 %
5.Claude Opus 5.558.6 %
6.Claude Sonnet 5.558.1 %
7.Gemini 3.5 Flash57.9 %
8.GLM-5.3-Flash57.9 %
9.Claude Fable 556.3 %
10.Gemini 3.6 Flash56.3 %
11.GLM-5.355.8 %
12.GPT-5.6 Luna55 %
13.GPT-5.6 Terra54.4 %
14.Kimi K354.4 %
15.Claude Opus 4.853.9 %
16.Claude Sonnet 553.9 %
17.GPT-5.6 Sol53.8 %
18.Grok 4.653.7 %
19.GPT-6 Astra53.5 %
20.DeepSeek-V4.1-Flash53.5 %
21.Grok 4.752.3 %
22.GPT-5.551.8 %
23.Claude Opus 4.751.5 %
24.Claude Sonnet 4.651 %
25.Qwen 3.8 Max 090250.6 %
26.GPT-6 Luna49.9 %
27.GLM-5.249.7 %
28.GPT-6 Sol49.1 %
29.Grok 4.548.3 %
30.Gemini 3.5 Flash-Lite47.4 %
31.GPT-5.4 mini45.4 %
32.Kimi K2.644.9 %
33.GLM-5.144.8 %
34.DeepSeek-V4-Pro44.1 %
35.Gemini 3.1 Pro Preview43 %
36.MiMo-V2.5-Pro41.5 %
37.Grok 4.337.7 %
38.MiMo-V2.536.7 %
39.Mistral Medium 3.532.1 %
Vals AIUpdated Sep 27, 2026 · 39 of 50 comparison models with an unambiguous result

A first-place result applies only to the named benchmark. It is not an overall quality score and says nothing by itself about price, speed, context length, or your own workload.

Decision criteria

What this LLM comparison lets you check

Benchmarks
Shared tasks and versions for reasoning, coding, agents, research, and other capabilities.
API pricing
Input, output, cache-read, and cache-write pricing where the provider publishes it.
Context
Published context windows and maximum output limits without imprecise rounding.
Architecture
Known parameters, active parameters, dense or MoE architecture, and open-weight status.
Availability
API access, lifecycle status, release date, and documented web-search support.
Sources
A retrieval or verification date and a direct source for each volatile result.

Need a broader directory? Browse the open-source LLM directory. For workload-specific cost estimates, use the API cost calculator.

Quick start

The 20 most popular model comparisons

Here you'll find the 20 matchups that matter right now. Compare general data, prices, benchmarks, and more.

  1. GPT-6.1 Sol vs. GPT-6 Sol
  2. GPT-6.1 Sol vs. Claude Sonnet 5.5
  3. Muse Spark 1.3 vs. Qwen 3.8 Max 0902
  4. GPT-5.6 Sol vs. Muse Spark 1.3
  5. Gemini 3.8 Flash vs. Muse Spark 1.3
  6. Muse Spark 1.3 vs. Claude Opus 5
  7. Qwen 3.8 Max 0902 vs. DeepSeek-V4-Pro
  8. Claude Fable 5.1 vs. Claude Mythos 5.1
  9. Claude Fable 5.1 vs. GPT-5.6 Sol
  10. Claude Opus 5.5 vs. Claude Fable 5.1
  11. Claude Opus 5 vs. GPT-6 Astra
  12. GPT-5.6 Sol vs. Gemini 3.8 Flash
  13. DeepSeek-V4-Pro vs. Kimi K3
  14. Grok 4.6 vs. Gemini 3.8 Flash
  15. GLM-5.2 vs. MiMo-V2.5-Pro
  16. Claude Opus 5 vs. Gemini 3.8 Flash
  17. Claude Opus 5 vs. Grok 4.6
  18. GPT-5.6 Sol vs. DeepSeek-V4-Pro
  19. Claude Sonnet 5.5 vs. GPT-5.6 Terra
  20. Claude Opus 5 vs. Kimi K3

Model selection

The 50 most relevant models

The selection includes current frontier models, lower-cost options, and selected earlier generations. You can compare Claude Opus 4.5 to 4.8, Sonnet 4.5 and 4.6, and Haiku 4.5 with newer models.

For the full technical profile of every model, browse the AI model directory.

Anthropic

Claude Opus 5.5Claude Sonnet 5.5Claude Fable 5.1Claude Mythos 5.1Claude Opus 5Claude Fable 5Claude Sonnet 5Claude Mythos 5Claude Opus 4.8Claude Opus 4.7Claude Opus 4.6Claude Opus 4.5Claude Sonnet 4.6Claude Sonnet 4.5Claude Haiku 4.5

OpenAI

GPT-6.1 SolGPT-6 AstraGPT-6 SolGPT-6 LunaGPT-5.6 SolGPT-5.6 TerraGPT-5.6 LunaGPT-5.5GPT-5.4GPT-5.4 mini

Google

Gemini 3.8 FlashGemini 3.7 FlashGemini 3.6 FlashGemini 3.5 FlashGemini 3.1 Pro PreviewGemini 3.5 Flash-Lite

Meta

Muse Spark 1.3

xAI

Grok 4.7Grok 4.6Grok 4.5Grok 4.3

DeepSeek

DeepSeek-V4-ProDeepSeek-V4-FlashDeepSeek-V4.1-Flash

Moonshot AI

Kimi K3Kimi K2.6

Z.ai

GLM-5.3-FlashGLM-5.3GLM-5.2GLM-5.1

Alibaba

Qwen 3.8 Max 0902

Mistral AI

Mistral Large 3Mistral Medium 3.5

Xiaomi

MiMo-V2.5MiMo-V2.5-Pro

Historical reference

A comparison from ChatGPT's early model generations

The current comparison matrix intentionally focuses on supported models. This historical reference keeps the documented differences between GPT-3.5 and GPT-4 available for existing applications and migration context.

GPT-3.5 vs. GPT-4: What's the Difference?

Frequently asked questions about the LLM comparison

How the benchmark matching, winners, and data freshness work.

There is no single best LLM for every workload. A coding agent, a low-cost extraction pipeline, and a long-context research task reward different strengths. Choose the two models you are considering above. The matchup page shows only shared measurement series and keeps benchmarks separate from pricing, context, architecture, and availability.

It compares large language models using shared benchmark results, pricing, technical specifications, availability, and source dates.

The comparison does not merge different benchmark versions, tasks, metrics, cohorts, reasoning levels, or agent setups into one score. A row appears only when both model results belong to the same documented measurement series.

No. A measured lead applies to the documented task and setup. Price, latency, context size, tool access, and your workload can change the practical choice.

The benchmark snapshot was retrieved on September 29, 2026. Volatile model facts also include their own source and verification date.

The database contains published result series for areas such as coding, reasoning, mathematics, knowledge, finance, law, agentic tasks, and multimodal capabilities. Individual leaderboards and matchup tables keep their documented task, metric, and cohort separate. The hub's orientation index averages rank percentiles only across setup-comparable cohorts, drawing on up to 10 of them; a model receives an overall rank once it is measured in at least 4 of those cohorts, scored as the mean percentile over the cohorts it appears in, and each row states how many it covers.

The first pricing table shows each developer's base rate per 1 million tokens. The matchup page then compares available provider endpoints separately. Regions, special rates, and differing endpoints are labeled so that unlike offers do not create a misleading price advantage.

Many providers publish neither the total nor the active parameter count. In that case, the table marks the value as not published. We do not estimate figures from rumors or older model generations.

The chart summarizes only directly comparable result series across five to eight capability categories. Results are normalized within each matched series, then averaged by benchmark family and category. The values describe the strengths in this matchup, not a universal model score.

Changelog

The latest updates and improvements to our LLM comparison
v1.12September 30, 2026

Claude Sonnet 5.5 added

  • Added Claude Sonnet 5.5 with official specifications, pricing, context window, knowledge cutoff, and source links
  • Expanded the comparison to 50 models and 1,225 head-to-head matchups, including GPT-6.1 Sol, Grok 4.7, and DeepSeek-V4.1-Flash
v1.11September 22, 2026

Claude Opus 5.5, GPT-6 Sol, and GPT-6 Luna added

  • Added Claude Opus 5.5, GPT-6 Sol, and GPT-6 Luna with official specifications, pricing, context windows, knowledge cutoffs, and source links
  • Expanded the comparison to 46 models and 1,035 head-to-head matchups
v1.10September 22, 2026

46-model roster and provider-balanced hosting prices

  • Expanded the comparison to 46 models and 1,035 head-to-head matchups
  • Switched hosting-market tariffs to the daily provider-balanced OpenRouter observation, so profiles and comparison pages show the same figure as the price index
  • Refreshed deprecation and lifecycle status against current provider pages
v1.9September 4, 2026

GPT-6 Astra added

  • Added GPT-6 Astra with official specifications, pricing, context window, availability, and source links
  • Expanded the comparison to 36 models and 630 head-to-head matchups
v1.8September 3, 2026

Muse Spark 1.3 and Qwen 3.8 Max 0902 added

  • Added Muse Spark 1.3 with current pricing, context, availability, and source links
  • Updated Qwen 3.8 Max to the 0902 snapshot and added it to the comparison roster
  • Expanded the comparison to 35 models and 595 head-to-head matchups
v1.7September 2, 2026

Gemini 3.8 Flash added

  • Added Gemini 3.8 Flash with official specifications, pricing, knowledge cutoff, and source links
  • Expanded the comparison to 33 models and 528 head-to-head matchups
v1.6September 2, 2026

Claude 5.1 and GLM-5.3 added

  • Added GLM-5.3 with official specifications, pricing, availability, and source links
  • Expanded the comparison to 32 models and 496 head-to-head matchups, including Claude Fable 5.1 and Mythos 5.1
v1.5August 29, 2026

GLM-5.3-Flash and cleaner leaderboards

  • Added GLM-5.3-Flash with official specifications, pricing, and source links
  • Expanded the comparison to 29 models and 406 head-to-head matchups
  • Removed models without a published score from leaderboard rows
v1.4August 28, 2026

Complete and cross-benchmark leaderboards

  • Added all 28 comparison models to every individual benchmark leaderboard
  • Added an orientation index from four setup-comparable benchmark cohorts
  • Excluded ambiguous model variants and kept missing results visible without estimation
v1.3August 24, 2026

Related matchups, provider icons, and expanded FAQ

  • Expanded each LLM matchup page to 14 related matchups and placed the hub button below the list
  • Added local provider icons to all 28 model badges without placeholders
  • Expanded the FAQ with benchmark, pricing, parameter, and radar methodology details
v1.2August 21, 2026

Benchmark leaderboards and capability profiles

  • Added five separate leaderboards for overall preference, reasoning, coding, agents, and finance
  • Added evidence-based capability radar charts to the head-to-head comparison pages
  • Kept every ranking within one published cohort, task, metric, and source snapshot
v1.1August 18, 2026

FAQ, changelog, and internal links

  • Added an FAQ about benchmark coverage, methodology, and database freshness
  • Added this changelog so material updates remain visible
  • Added the hub and all matchup pages to navigation, the sitemap, and related comparisons
v1.0August 17, 2026

Initial release

  • Published 28 current models and 378 head-to-head matchups
  • Combined shared benchmarks, API pricing, specifications, and primary sources
  • Added the comparison picker and ten prioritized matchups