Text generation and understanding
44
Models
27
Benchmarks
13
Providers
29
Open Weight
43
Deployable Now
28 self-host
44
Structured benchmark-backed rows
0
Signal-backed rows without full normalized tables
0
Benchmark-expected rows still catching up
| # | Model | Capability | Quality | Params | MATH Lvl 5 | HumanEval | MMLU | GPQA | $/M In | Speed | Open |
|---|---|---|---|---|---|---|---|---|---|---|---|
97 | grok-4 xAI Subscription |
| 87.1 |
| 67.0 |
| Undisclosed |
| 99.0 |
| 81.9 |
| 86.6 |
| 87.7 |
$3.00/M Cheapest verified |
| — |
1 | GPT-5.5 OpenAI SubscriptionStructured | 85.2 | 65.8 | Undisclosed | — | — | — | 93.6 | $5.00/M Cheapest verified | — |
55 | DeepSeek-V4-Pro DeepSeek Ready to UseStructured | 82.1 | 68.5 | Unknown | — | 76.8 | 90.8 | 90.1 | Free Free | — |
41 | DeepSeek-R1 DeepSeek Self-HostStructured | 81.9 | 80.4 | 671B | 97.3 | — | 90.8 | 71.5 | $0.2800/M Cheapest verified | — |
4 | GPT-5.4 OpenAI SubscriptionStructured | 81.2 | 68.3 | Undisclosed | — | — | — | 92.8 | $1.25/M Cheapest verified | — |
51 | GPT-4.5 OpenAI SubscriptionStructuredArchived | 80.3 | 60.4 | Undisclosed | — | — | — | — | $75.00/M Cheapest verified | — |
9 | GPT-5.2 OpenAI SubscriptionStructured | 80.2 | 59.3 | Undisclosed | — | — | 89.6 | 92.4 | $1.25/M Cheapest verified | — |
14 | DeepSeek V3.2 Speciale DeepSeek Open WeightsStructured | 79.8 | 64.7 | Unknown | — | 89.6 | 86.3 | 87.1 | $0.2800/M Cheapest verified | — |
16 | GPT-5.1 OpenAI SubscriptionStructured | 77.9 | 56.3 | Undisclosed | — | — | — | 88.1 | $1.25/M Cheapest verified | — |
39 | Qwen3-235B Alibaba Open WeightsStructured | 77.2 | 79.6 | 235B | — | — | — | — | $0.8000/M Cheapest verified | 72 tok/s |
35 | o3 OpenAI API AccessStructured | 76.9 | 59.7 | Undisclosed | 99.2 | 80.8 | 85.3 | 84.5 | $20.00/M Cheapest verified | — |
31 | DeepSeek-V3 DeepSeek Open WeightsStructured | 76.7 | 79.6 | 685B | 90.2 | — | 87.1 | 59.1 | $0.2800/M Cheapest verified | 145 tok/s |
18 | MiniMax M2.7 MiniMax SubscriptionStructured | 76.3 | 42.0 | Undisclosed | — | 47.0 | — | 87.4 | $0.2790/M Cheapest verified | — |
24 | MiniMax M2.5 MiniMax Open WeightsStructured | 75.5 | 47.9 | Unknown | — | 42.6 | — | 84.8 | $0.1500/M Cheapest verified | — |
27 | GPT-4.1 OpenAI SubscriptionStructured | 74.9 | 60.6 | Undisclosed | 91.3 | 45.7 | 80.6 | 66.6 | $2.00/M Cheapest verified | — |
15 | MiniMax M2.1 MiniMax SubscriptionStructured | 74.7 | 44.1 | Undisclosed | — | 81.0 | 87.5 | 83.0 | $0.2900/M Cheapest verified | — |
29 | Grok 4.3 xAI SubscriptionStructured | 74.4 | 58.5 | Undisclosed | — | 47.3 | — | 90.1 | $1.25/M Cheapest verified | — |
43 | Llama 4 Maverick Meta Open WeightsStructured | 73.6 | 72.9 | 400B | 88.9 | 39.7 | 80.9 | 67.1 | Free Cheapest verified | 88 tok/s |
136 | qwen3-235b-a22b-instruct-2507 Qwen Open WeightsStructured | 72.9 | 72.5 | 235B | — | — | — | — | Free Free | — |
40 | Qwen3 Max Thinking Qwen Open WeightsStructured | 72.9 | 42.2 | Unknown | — | 43.1 | — | 86.1 | $0.7800/M Cheapest verified | — |
47 | GPT-5 OpenAI SubscriptionStructured | 72.1 | 50.9 | Undisclosed | — | — | — | 85.7 | $1.25/M Cheapest verified | — |
36 | Kimi K2 Moonshot AI SubscriptionStructured | 71.6 | 55.5 | Undisclosed | 97.1 | 55.6 | 82.4 | 76.6 | $0.5700/M Cheapest verified | — |
26 | MiniMax M2 MiniMax Open WeightsStructured | 71.5 | 47.9 | Unknown | — | 82.6 | 82.0 | 77.7 | $0.2550/M Cheapest verified | — |
94 | llama-3.1-405b-instruct Meta Open WeightsStructured | 70.2 | 64.3 | 405B | — | 89.0 | 87.3 | — | Free Cheapest verified | — |
17 | QwQ 32B Qwen Open WeightsStructured | 69.8 | 64.1 | 32B | 95.7 | 63.1 | 76.4 | 59.3 | $0.1500/M Cheapest verified | — |
87 | Nemotron 3 Super NVIDIA Open WeightsStructured | 69.5 | 43.4 | 120B | — | — | — | 79.2 | $0.0900/M Cheapest verified | — |
37 | Qwen3 Max Qwen Open WeightsStructured | 69.1 | 43.6 | Unknown | — | 76.7 | 84.1 | 76.4 | $0.7800/M Cheapest verified | — |
116 | deepseek-v3.1 DeepSeek Open WeightsStructured | 68.1 | 71.4 | Unknown | — | — | — | — | $0.2800/M Cheapest verified | — |
162 | Mistral Large 2 Mistral AI Structured | 68.0 | 57.9 | 123B | — | — | — | — | $2.00/M Cheapest verified | — |
137 | Nemotron 70B NVIDIA Open WeightsStructured | 64.9 | 65.5 | 70B | — | — | — | — | Free Cheapest verified | — |
45 | GPT-4 Turbo OpenAI SubscriptionStructured | 63.9 | 51.0 | Undisclosed | 73.7 | 29.1 | 69.4 | — | $10.00/M Cheapest verified | — |
128 | DeepSeek-V3.2 DeepSeek Open WeightsStructured | 63.6 | 72.6 | Unknown | — | — | — | — | $0.2800/M Cheapest verified | — |
72 | Llama 3.3 70B Meta Open WeightsStructured | 61.4 | 68.3 | 70B | 48.3 | — | 86.0 | 10.5 | $0.5400/M Cheapest verified | — |
131 | Command R+ Cohere Open WeightsStructured | 59.3 | 63.6 | 104B | — | — | 75.7 | — | $2.50/M Cheapest verified | — |
23 | Phi 4 Microsoft Open WeightsStructured | 59.2 | 55.2 | 14B | 50.0 | 23.1 | 71.4 | 11.5 | $0.0650/M Cheapest verified | — |
63 | Gemma 3 27B | 58.7 | 63.5 | 27B | — | — | — | — | Free Cheapest verified | — |
82 | Qwen2.5-72B Alibaba Open WeightsStructured | 57.9 | 67.4 | 72B | 39.1 | — | 86.1 | 20.7 | $0.6000/M Cheapest verified | — |
93 | llama-3-70b-instruct Meta Open WeightsStructured | 56.8 | 63.4 | 70B | — | 81.7 | 82.0 | 39.5 | Free Free | — |
62 | mistral-large Mistral AI Open WeightsStructured | 51.3 | 46.1 | Unknown | — | — | 84.0 | 46.3 | $2.00/M Cheapest verified | — |
52 | Llama 3.1 Nemotron 70B Instruct NVIDIA Open WeightsStructured | 47.9 | 51.4 | 70B | 42.7 | — | — | 1.1 | $1.20/M Cheapest verified | — |
120 | Mixtral 8x22B Instruct Mistral AI Open WeightsStructured | 38.9 | 33.0 | 176B | 54.5 | 14.8 | 53.7 | 33.2 | $0.5000/M Cheapest verified | — |
223 | Qwen3-0.6B Qwen Open WeightsStructured | 30.2 | 45.5 | 600M | — | — | — | — | Free Free | — |
204 | mixtral-8x7b-instruct-v0.1 Mistral AI Open WeightsStructured | 27.0 | 43.7 | 56B | — | — | — | — | Free Free | — |
207 | mistral-7b-instruct-v0.2 Mistral AI Open WeightsStructured | 15.5 | 37.7 | 7B | — | — | — | — | Free Free | — |