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
451Nova Pro 1.0Amazon———300K$1.40
452Amazon Nova ProAmazon123.8————
453Amazon Nova LiteAmazon—————
454Amazon Nova MicroAmazon—————
455Yi-Lightning01.AI—————
456INTELLECT-1Prime Intellect,Hugging Face,Arcee AI100.4————
457Tulu 3 (Tülu 3) 70BAllen Institute for AI—46.3———
458GPT-4o (Nov 2024)OpenAI128.847.9———
459GPT-4o (2024-11-20)OpenAI———128K$4.38
460Mistral Large 2407Mistral AI———131K$3.00
461Mistral Large 2 (Nov 2024)Mistral AI128.551.3———
462Qwen2.5 Coder 32B InstructAlibaba (Qwen)———33K$0.74
463UnslopNemo 12BTheDrummer———1.02M$0.40
464Qwen2.5-Coder-14B-Instruct—————
465Qwen2.5-Coder-3B-Instruct—————
466Qwen-TurboAlibaba (Qwen)—41.8———
467Claude 3.5 Sonnet (October 2024)Anthropic133.6————
468Claude 3.5 HaikuAnthropic127.2————
469Claude 3.5 Haiku (Oct 2024)Anthropic—38.1———
470Claude 3.5 Sonnet (Oct 2024)Anthropic—55.3———
471Magnum v4 72Banthracite-org———33K$3.13
472Ministral 3BMistral AI118.125.3———
473Ministral 8BMistral AI—27.1———
474Qwen2.5 7B InstructAlibaba (Qwen)———33K$0.13
475Gemini 1.5 Flash 8BGoogle—33.0———
476Llama 3.2 1B InstructMeta———60K$0.070
477Llama 3.2 3B InstructMeta———131K$0.12
478Gemini 1.5 Pro (Sept 2024)Google131.757.2———
479Gemini 1.5 Flash (Sep 2024)Google129.447.3———
480Llama 3.2 90BMeta125.541.0———
481Llama 3.2 1BMeta102.023.9———
482Llama 3.2 11BMeta—————
483Llama 3.2 3BMeta—————
484Qwen2.5-72BAlibaba (Qwen)129.049.1———
485Qwen2.5-7BAlibaba (Qwen)118.435.5———
486Qwen2.5 72B InstructAlibaba (Qwen)———33K$0.37
487Qwen2.5-14BAlibaba (Qwen)—————
488Qwen2.5-Coder-32BAlibaba (Qwen)119.4————
489Qwen2.5-Coder-14B116.3————
490Qwen2.5-Coder (7B)Alibaba (Qwen)112.9————
491Qwen2.5-Coder-3B107.4————
492Qwen2.5-Coder (1.5B)Alibaba (Qwen)102.6————
493Qwen2.5-Coder-0.5B87.4————
494Qwen2.5-32BAlibaba (Qwen)128.546.1———
495Mistral Small v24.09Mistral AI—————
496Pixtral 12BMistral AI—————
497o1-miniOpenAI · +2 variants135.862.4———
498o1-previewOpenAI134.850.3———
499DeepSeek-V2.5 (Sep 2024)DeepSeek—————
500Command R+Cohere119.3————
— 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