We put the same work to every large language model. Then we ranked them.
One identical set of tasks, 73 models, scored on craft, speed, control and value. Closed APIs and open weights on the same page, because that is the choice.
- Tracked
- 73
- Fully reviewed
- 00
- Categories
- 05
- Average score
- 81.1
The short answer
What is the best AI model in 2026?
Claude Opus 5.5 leads our index of 73 large language models with 92.8 out of 100, ahead of Claude Sonnet 5.5 on 91.5 and Gemini 4 Argon on 90.6. Models are scored on answer quality, latency, steerability and capability per dollar, with answer quality weighted heaviest. Nothing here is sponsored.
Index updated
The model index
Every large language model we track, in rank order.
Frontier flagships, open weights you can host yourself, and small models quick enough to sit in a loop. Narrow by what kind of model it is, or re-sort by what you care about.
73 builders
0192.8Claude Opus 5.5
Fable-level answers for less than the Opus it replaces. The new default.
FrontierProvisional
- 0291.5
Claude Sonnet 5.5
Nearly the strongest model in the index, at a mid-range price. Sonnet stopped being the compromise.
CodingProvisional
- Gemini 4 Argon0390.6
Gemini 4 Argon
Top of the human-preference leaderboard, and almost nobody can use it yet.
FrontierProvisional
0490.3GPT-6.1 Sol
Near-Astra answers at a fifth of the price. The OpenAI model most teams should deploy now.
FrontierProvisional
- GLM-5.3 Flash0589.6
GLM-5.3 Flash
A fast, MIT-licensed open model that human voters rank beside the big flagships.
Open weightsProvisional
- MiMo V2.6 Pro0689.4
Xiaomi MiMo V2.6 Pro
The strongest large open-weights model on the independent index, under MIT, for under a dollar out.
Open weightsProvisional
- DeepSeek V4.1 Flash0788.3
DeepSeek V4.1 Flash
Frontier-adjacent answers for thirty cents per million input tokens, with the weights under MIT.
Open weightsProvisional
- 0887.9
Claude Fable 5.1
Anthropic's top tier for the longest, hardest runs, with a cheaper Opus now snapping at its heels.
FrontierProvisional
- Gemini 3.8 Flash0987.9
Gemini 3.8 Flash
Google's smartest Flash yet, eighth on the arena, and still free to start on.
FrontierProvisional
- GLM-5.31087.9
GLM-5.3
The same base as GLM-5.2, re-trained into one of the strongest open models available.
Open weightsProvisional
- 1187.9
Claude Haiku 5.5
A small model priced like a rounding error that scores where mid-tier models did last quarter.
Small and fastProvisional
- 1287.1
Claude Opus 5
The default of the summer, now replaced by an Opus that is cheaper and better.
FrontierProvisional
- GLM-5.21387.1
GLM-5.2
Still the GLM to choose when the licence matters most: MIT, a million tokens, no conditions.
Open weightsProvisional
- MiMo V2.6 Flash1486.9
Xiaomi MiMo V2.6 Flash
Xiaomi's fast tier: a sparse open model at fourteen cents in and twenty-eight out.
Open weightsProvisional
- Qwen 3.8 Flash1586.5
Qwen 3.8 Flash
A preview of the architecture behind Qwen 4, open, and priced at fifteen cents in.
Open weightsProvisional
The complete index
Every entry on this page, in rank order. The cards above load in batches; this list does not.
- 01Claude Opus 5.592.8
- 02Claude Sonnet 5.591.5
- 03Gemini 4 Argon90.6
- 04GPT-6.1 Sol90.3
- 05GLM-5.3 Flash89.6
- 06Xiaomi MiMo V2.6 Pro89.4
- 07DeepSeek V4.1 Flash88.3
- 08Claude Fable 5.187.9
- 09Gemini 3.8 Flash87.9
- 10GLM-5.387.9
- 11Claude Haiku 5.587.9
- 12Claude Opus 587.1
- 13GLM-5.287.1
- 14Xiaomi MiMo V2.6 Flash86.9
- 15Qwen 3.8 Flash86.5
- 16Muse Spark 1.385.7
- 17Claude Fable 585.6
- 18Qwen 3.8-Max85.1
- 19Gemini 3.6 Flash85.1
- 20Kimi K385.1
- 21GPT-6 Astra84.9
- 22GPT-6 Luna84.9
- 23Grok 4.784.5
- 24MiniMax M384.5
- 25GPT-5.6 Sol84.4
- 26Hunyuan Hy4 Preview84.2
- 27GPT-5.6 Terra83.6
- 28DeepSeek V4 Pro83.4
- 29Step 5 Preview83.2
- 30GPT-5.6 Luna83.0
- 31Qwen 3.8 27B82.9
- 32Doubao Seed 2.1 Pro82.6
- 33Claude Sonnet 582.5
- 34Qwen 3.682.2
- 35Mistral Large 482.1
- 36Qwen 3.5 397B81.9
- 37Step 3.7 Flash81.6
- 38Doubao Seed 2.0 Pro81.5
- 39Grok 4.581.3
- 40Mistral Large 380.7
- 41Llama 4 Maverick80.4
- 42Qwen 3.7-Max80.2
- 43Gemma 480.1
- 44Hunyuan HY380.0
- 45Solar Open 279.2
- 46Hermes 4.378.9
- 47Muse Glimmer 30B78.4
- 48Solar Pro 478.2
- 49Mistral Small 478.1
- 50Mistral Medium 3.578.0
- 51Nemotron 3 Ultra78.0
- 52Nemotron 3 Super77.9
- 53K-EXAONE 2.077.8
- 54ERNIE 5.177.7
- 55openPangu 2.0 Pro77.7
- 56Llama 4 Scout77.3
- 57Grok 4.2076.9
- 58Cohere Command A+76.8
- 59Phi-4 Reasoning Vision76.8
- 60IBM Granite 4.276.4
- 61Microsoft MAI-Thinking-176.0
- 62Mercury 2.575.2
- 63Gemini 3.5 Flash-Lite75.2
- 64Amazon Nova 2 Lite74.5
- 65Sarvam 105B74.3
- 66OLMo 3.172.0
- 67Doubao Seed 1.6 Flash71.2
- 68Claude Haiku 4.571.2
- 69HyperCLOVA X THINK70.7
- 70Ministral 369.0
- 71Liquid LFM 2.568.9
- 72Reka Edge68.7
- 73SmolLM367.7
Weekly update
The ranking moves. Find out when.
One short email a week: what changed in the index, which builder moved and why, and anything new worth trying. No sponsorships, no reprinted press releases.