Skip to content

Newest models

Most recent releases, with whatever independent scores exist so far.

Most recent releases, with whatever independent scores exist so far. Source: OpenRouter / Epoch AI. Variants (reasoning effort, batch) are folded into one row.

#ModelCapabilityReasoningCodingContext$/M
351Gemini 2.5 Pro (Jun 2025)Google145.385.3———
352Gemini 2.5 Pro Preview (Jun 2025)Google—84.8———
353Gemini 2.5 Pro Preview 06-05Google———1.05M$3.44
354DeepSeek-R1 (May 2025)DeepSeek141.376.3———
355deepseek-r1-0528-qwen3-8b—9.3———
356R1 0528DeepSeek———164K$0.91
357Claude Opus 4Anthropic142.776.3———
358Claude Sonnet 4Anthropic · +1 variant141.779.2—200K$6.00
359Gemini 2.5 Flash (May 2025)Google141.5————
360codex-mini-2025-05-16—————
361Mistral Medium 3Mistral AI134.159.5—131K$0.80
362Gemini 2.5 Pro (May 2025)Google142.5————
363Llama Guard 4 12BMeta———164K$0.18
364Qwen3 32BAlibaba (Qwen)138.565.7—131K$0.13
365Qwen3 14BAlibaba (Qwen)138.263.8—131K$0.15
366Qwen3 30B A3BAlibaba (Qwen)136.261.7—131K$0.23
367Qwen3-1.7BAlibaba (Qwen)—38.0———
368Qwen3-4BAlibaba (Qwen)—52.3———
369Qwen3 235B A22BAlibaba (Qwen)139.470.7—131K$0.80
370Qwen3 8BAlibaba (Qwen)136.256.8—131K$0.20
371Qwen Plus (Apr 2025)Alibaba (Qwen)—————
372Gemini 2.5 Flash (Apr 2025)Google140.0————
373Gemini 2.5 Flash Preview (Apr 2025)Google—————
374o3OpenAI · +4 variants146.981.8—200K$3.50
375o4 MiniOpenAI · +1 variant145.679.6—200K$1.93
376o4 Mini HighOpenAI———200K$1.93
377o4-minimediumOpenAI · +1 variant—77.8———
378GPT-4.1OpenAI · +1 variant136.866.9—1.05M$3.50
379GPT-4.1 MiniOpenAI · +1 variant135.065.8—1.05M$0.70
380GPT-4.1 NanoOpenAI · +1 variant129.648.9—1.05M$0.17
381Grok 3xAI138.375.8———
382QWQ-PlusAlibaba (Qwen)—65.4———
383Llama 4 MaverickMeta132.2——1.05M$0.30
384Llama 4 ScoutMeta129.651.8—1.31M$0.15
385Llama 4 Maverick (FP8)Meta—67.0———
386Gemini 2.5 Pro (Mar 2025)Google144.2————
387Gemini 2.5 Pro Preview (Mar 2025)Google—————
388GPT-4o (Mar 2025)OpenAI—————
389Gemini 2.5 Pro Exp (Mar 2025)Google—83.8———
390DeepSeek-V3 (Mar 2025)DeepSeek135.967.6———
391DeepSeek V3 0324DeepSeek———164K$0.50
392o1 ProlowOpenAI—————
393o1-proOpenAI———200K$262.50
394Mistral Small 3.1Mistral AI127.547.5———
395Mistral Small 3.1 24BMistral AI———128K$0.40
396Cohere Command ACohere—————
397Command ACohere———256K$4.38
398Gemma 3 27BGoogle130.047.7—131K$0.17
399Gemma 3 12BGoogle123.539.5—131K$0.075
400Gemma 3 4BGoogle116.023.2—131K$0.063
— means no published score from that source yet · click a model for every benchmark with its source
Best AI for coding →
Models ranked on DeepSWE
Best AI for reasoning →
Ranked on GPQA Diamond
Best AI for math →
Ranked on competition math
Best open-weight models →
Weights you can run yourself
Cheapest capable models →
Lowest blended API price
Best AI coding tools →
Apps & agents by popularity
Best AI image generators →
Apps & models
Best AI for research →
Answer engines & deep research