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
151Qwen3 32BAlibaba (Qwen)138.565.7—131K$0.13
152Qwen3 14BAlibaba (Qwen)138.263.8—131K$0.15
153Qwen3 30B A3BAlibaba (Qwen)136.261.7—131K$0.23
154Qwen3-1.7BAlibaba (Qwen)—38.0———
155Qwen3-4BAlibaba (Qwen)—52.3———
156Qwen3 235B A22BAlibaba (Qwen)139.470.7—131K$0.80
157Qwen3 8BAlibaba (Qwen)136.256.8—131K$0.20
158Llama 4 MaverickMeta132.2——1.05M$0.30
159Llama 4 ScoutMeta129.651.8—1.31M$0.15
160Llama 4 Maverick (FP8)Meta—67.0———
161DeepSeek-V3 (Mar 2025)DeepSeek135.967.6———
162DeepSeek V3 0324DeepSeek———164K$0.50
163Mistral Small 3.1Mistral AI127.547.5———
164Mistral Small 3.1 24BMistral AI———128K$0.40
165Cohere Command ACohere—————
166Command ACohere———256K$4.38
167Gemma 3 27BGoogle130.047.7—131K$0.17
168Gemma 3 12BGoogle123.539.5—131K$0.075
169Gemma 3 4BGoogle116.023.2—131K$0.063
170Gemma 3 1BGoogle—19.9———
171Reka Flash 3Reka AI———66K$0.13
172Skyfall 36B V2TheDrummer———33K$0.61
173QwQ-32BAlibaba (Qwen)137.665.3———
174Qwen2.5 VL 72B InstructAlibaba (Qwen)———128K$0.85
175Mistral Small 3Mistral AI127.147.3—33K$0.058
176R1DeepSeek139.071.7—64K$1.15
177DeepSeek-R1-Distill-Qwen-32BDeepSeek137.464.1———
178DeepSeek-R1-Distill-Qwen-14BDeepSeek135.444.7———
179DeepSeek-R1-Distill-Llama-70BDeepSeek—55.7———
180DeepSeek-R1-Distill-Qwen-1.5BDeepSeek—33.6———
181MiniMax-01MiniMax———1M$0.42
182CodestralMistral AI—————
183Eurus-2-7B-PRIMETsinghua University,University of Illinois Urbana-Champaign (UIUC),Shanghai AI Lab,Peking University,Shanghai Jiao Tong University,CUHK Shenzhen Research Institute—33.9———
184DeepSeek V3DeepSeek132.456.5—164K$0.45
185Llama 3.3 Euryale 70BSao10K———131K$0.68
186Phi 4Microsoft130.456.1—16K$0.087
187Llama 3.3 70BMeta127.347.4———
188Llama 3.3 70B InstructMeta———131K$0.15
189Tulu 3 (Tülu 3) 70BAllen Institute for AI—46.3———
190Mistral Large 2 (Nov 2024)Mistral AI128.551.3———
191Qwen2.5 Coder 32B InstructAlibaba (Qwen)———33K$0.74
192UnslopNemo 12BTheDrummer———1.02M$0.40
193Magnum v4 72Banthracite-org———33K$3.13
194Ministral 8BMistral AI—27.1———
195Qwen2.5 7B InstructAlibaba (Qwen)———33K$0.13
196Llama 3.2 1B InstructMeta———60K$0.070
197Llama 3.2 3B InstructMeta———131K$0.12
198Llama 3.2 90BMeta125.541.0———
199Llama 3.2 1BMeta102.023.9———
200Llama 3.2 11BMeta—————
— 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