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
201Qwen3 Coder NextAlibaba (Qwen)———262K$0.29
202Step 3.5 FlashStepFun———262K$0.15
203Kimi K2.5Moonshot AI148.1——262K$0.90
204Kimi K2.5 (Fireworks)Moonshot AI—87.6———
205Solar Pro 3Upstage———131K$0.26
206MiniMax M2-herMiniMax———66K$0.53
207Palmyra X5Writer———1.04M$1.95
208GLM 4.7 FlashZ.ai (Zhipu AI)—60.5—200K$0.15
209GPT AudioOpenAI———128K$4.38
210GPT Audio MiniOpenAI———128K$1.05
211GPT-5.2-CodexOpenAI———400K$4.81
212MiniMax M2.1MiniMax———205K$0.53
213Seed 1.6ByteDance———262K$0.69
214Seed 1.6 FlashByteDance———262K$0.13
215GLM 4.7Z.ai (Zhipu AI)143.583.3—205K$1.00
216GLM-4.7 (Novita)Z.ai (Zhipu AI)—————
217GLM-4.7 (Together)Z.ai (Zhipu AI)—————
218Gemini 3 FlashGoogle151.8————
219Gemini 3 Flash PreviewGoogle · +1 variant—89.4—1.05M$1.13
220Nemotron 3 Nano 30B A3BNVIDIA———262K$0.087
221GPT-5.2 ProOpenAI · +1 variant155.4——400K$57.75
222GPT-5.2OpenAI · +6 variants153.488.2—400K$4.81
223GPT-5.2 Pro (web)OpenAI—————
224GPT-5.2 ChatOpenAI———128K$4.81
225Devstral 2 2512Mistral AI———262K$0.80
226GLM 4.6VZ.ai (Zhipu AI)———131K$0.45
227Relace SearchRelace———256K$1.50
228GPT-5.1-Codex-MaxOpenAI———400K$3.44
229Ministral 3 14B 2512Mistral AI———262K$0.20
230Ministral 3 3B 2512Mistral AI———131K$0.10
231Ministral 3 8B 2512Mistral AI · +1 variant———262K$0.15
232Mistral Large 3Mistral AI—————
233Nova 2 LiteAmazon———1M$0.85
234DeepSeek V3.2DeepSeek146.383.4—164K$0.32
235DeepSeek-V3.2 (Thinking; Fireworks)DeepSeek—————
236DeepSeek-V3.2 (Thinking; Novita)DeepSeek—————
237DeepSeek-V3.2-SpecialeDeepSeek—————
238Mistral Large 3 2512Mistral AI · +1 variant———262K$0.75
239Claude Opus 4.5Anthropic · +10 variants150.1——200K$10.00
240Nano Banana Pro (Gemini 3 Pro Image Preview)Google———66K$4.50
241Grok 4.1 FastxAI—————
242Gemini 3 ProGoogle153.0————
243Gemini 3 Pro PreviewGoogle · +1 variant—92.6———
244Grok 4.1xAI—————
245GPT-5.1OpenAI · +5 variants149.787.6—400K$3.44
246GPT-5.1-CodexOpenAI———400K$3.44
247GPT-5.1-Codex-MiniOpenAI———400K$0.69
248Kimi K2 ThinkingMoonshot AI146.0——262K$1.07
249Kimi K2 Thinking (Together)Moonshot AI—————
250Kimi K2 Thinking TurboMoonshot AI—84.2———
— 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