Text generation and understanding
43
Models
27
Benchmarks
12
Providers
28
Open Weight
43
Deployable Now
27 self-host
43
Structured benchmark-backed rows
0
Signal-backed rows without full normalized tables
0
Benchmark-expected rows still catching up
| # | Model | Economic Footprint | Quality | Params | TAU-Bench | OSWorld | GPQA | GAIA | $/M In | Speed | Open |
|---|---|---|---|---|---|---|---|---|---|---|---|
2 | GPT-5.6 Sol OpenAI Subscription |
| 74.0 |
| 71.8 |
| Undisclosed |
| 46.9 |
| 62.6 |
| 94.6 |
| 77.4 |
$2.00/M Cheapest verified |
| — |
73 | GPT-5.6 Terra OpenAI SubscriptionStructured | 67.8 | 69.5 | Undisclosed | — | 62.6 | 94.6 | — | $2.00/M Cheapest verified | — |
19 | Llama 3.1 8B Instruct Meta Open WeightsStructured | 63.9 | 61.7 | 8B | — | — | 39.5 | 0.7 | Free Cheapest verified | — |
69 | GPT-4 Turbo OpenAI SubscriptionStructured | 62.8 | 37.8 | Undisclosed | — | — | — | 40.0 | $10.00/M Cheapest verified | — |
10 | GPT-5.6 Luna OpenAI SubscriptionStructured | 62.7 | 71.0 | Undisclosed | — | 62.6 | 94.6 | — | $0.2000/M Cheapest verified | — |
85 | Llama 3.3 70B Instruct Meta Open WeightsStructured | 61.3 | 60.7 | 70B | — | — | 10.5 | — | $0.1000/M Cheapest verified | — |
97 | Phi 4 Microsoft Open WeightsStructured | 59.3 | 59.2 | 15B | — | — | 11.5 | — | $0.0700/M Cheapest verified | — |
98 | Qwen3 235B A22B Qwen Open WeightsStructured | 58.3 | 70.6 | 235B | — | — | — | 44.2 | $0.4550/M Cheapest verified | — |
71 | o3 OpenAI API AccessStructured | 55.6 | 48.2 | Undisclosed | 82.5 | 45.2 | 82.7 | 30.2 | $2.00/M Cheapest verified | — |
65 | GPT-4.1 OpenAI SubscriptionStructured | 55.6 | 46.0 | Undisclosed | 71.8 | 32.5 | 66.6 | 71.4 | $2.00/M Cheapest verified | — |
252 | Kimi K2 Moonshot AI SubscriptionStructured | 55.0 | 72.2 | 1.0T | — | — | 76.6 | — | $0.6600/M Cheapest verified | — |
130 | Qwen3 235B A22B Thinking 2507 Qwen Open WeightsStructured | 53.4 | 67.4 | 235B | — | — | — | 44.2 | $0.2300/M Cheapest verified | — |
9 | GPT-5.4 OpenAI SubscriptionStructured | 53.3 | 57.1 | Undisclosed | 98.9 | 75.0 | 92.8 | 68.1 | $1.25/M Cheapest verified | — |
3 | GPT-5.2 OpenAI SubscriptionStructured | 53.1 | 59.3 | Undisclosed | 98.7 | — | 92.4 | 25.2 | $1.25/M Cheapest verified | — |
24 | Command A Cohere Open WeightsStructured | 53.1 | 45.9 | Unknown | — | — | 76.1 | — | $2.50/M Cheapest verified | — |
81 | Qwen3.8 2.4T A95B Qwen Self-HostStructured | 52.7 | 65.0 | 2.4T | — | — | 93.5 | 44.2 | $2.00/M Cheapest verified | — |
108 | Qwen3.5 397B A17B Qwen Open WeightsStructured | 52.6 | 62.5 | 403B | 68.4 | — | — | 44.2 | $0.5500/M Cheapest verified | — |
12 | GPT-5.1 OpenAI SubscriptionStructured | 52.5 | 55.9 | Undisclosed | 95.6 | — | 88.1 | 22.9 | $1.25/M Cheapest verified | — |
72 | GPT-5 OpenAI SubscriptionStructured | 51.8 | 41.3 | Undisclosed | 96.7 | 62.6 | 94.6 | 0.7 | $1.25/M Cheapest verified | — |
158 | Kimi K2 Thinking Moonshot AI SubscriptionStructured | 49.9 | 48.5 | 1.0T | — | — | 83.8 | — | $0.6000/M Cheapest verified | — |
181 | Llama 3.2 3B Instruct Meta Open WeightsStructured | 49.4 | 46.9 | 3B | — | — | — | — | $0.0500/M Cheapest verified | — |
110 | Qwen2.5-72B Alibaba Open WeightsStructured | 49.3 | 63.1 | 72B | — | — | 20.7 | 4.7 | $0.6000/M Cheapest verified | — |
290 | Llama 3 70B Instruct Meta Open WeightsStructured | 49.1 | 58.1 | 70B | — | — | 39.5 | 14.0 | $0.5100/M Cheapest verified | — |
79 | DeepSeek-R1 DeepSeek Self-HostStructured | 48.7 | 76.9 | 671B | — | — | 71.5 | 13.0 | $0.2800/M Cheapest verified | — |
107 | Qwen3 Coder Next Qwen Open WeightsStructured | 48.7 | 49.2 | 80B | — | — | 73.7 | 44.2 | $0.1200/M Cheapest verified | — |
89 | DeepSeek-V3 DeepSeek Open WeightsStructured | 48.5 | 73.4 | 685B | — | — | 59.1 | 13.0 | $0.2800/M Cheapest verified | 145 tok/s |
80 | Llama 4 Maverick Meta Open WeightsStructured | 47.3 | 70.2 | 400B | 45.7 | 15.3 | 67.1 | — | Free Cheapest verified | 88 tok/s |
261 | Nemotron 70B NVIDIA Open WeightsStructured | 47.2 | 63.4 | 70B | — | — | — | — | Free Cheapest verified | — |
21 | DeepSeek V3.1 Terminus DeepSeek Open WeightsStructured | 45.1 | 46.9 | 671B | — | — | — | 13.0 | $0.2700/M Cheapest verified | — |
182 | Grok-2 xAI SubscriptionStructured | 45.0 | 41.4 | Undisclosed | — | — | — | — | $2.00/M Cheapest verified | — |
248 | grok-4 xAI SubscriptionStructured | 44.5 | 66.9 | 3.0T | 18.0 | — | 87.7 | — | $3.00/M Cheapest verified | — |
95 | Qwen3-235B Alibaba Open WeightsStructured | 44.3 | 72.5 | 235B | 48.1 | 14.7 | — | 44.2 | $0.8000/M Cheapest verified | 72 tok/s |
426 | QwQ 32B Qwen Open WeightsStructured | 44.3 | 70.3 | 33B | — | — | 59.3 | 40.6 | $0.1500/M Cheapest verified | — |
51 | Qwen3 Max Thinking Qwen Open WeightsStructured | 40.8 | 25.7 | Unknown | — | — | 86.1 | 44.2 | $0.7800/M Cheapest verified | — |
16 | MiniMax M2.5 MiniMax Open WeightsStructured | 40.2 | 46.0 | Unknown | — | — | 84.8 | — | $0.2700/M Cheapest verified | — |
31 | Qwen3.7 Max Qwen Open WeightsStructured | 39.3 | 30.6 | Unknown | — | — | 92.3 | 44.2 | $1.48/M Cheapest verified | — |
331 | qwen3-235b-a22b-instruct-2507 Qwen Open WeightsStructured | 37.1 | 71.8 | 235B | — | — | — | 44.2 | Free Free | — |
27 | MiniMax M2.1 MiniMax SubscriptionStructured | 37.1 | 26.0 | Undisclosed | — | — | 83.0 | — | $0.3000/M Cheapest verified | — |
287 | DeepSeek-V3.2 DeepSeek Open WeightsStructured | 35.4 | 68.0 | Unknown | — | 18.9 | — | 13.0 | $0.2800/M Cheapest verified | — |
91 | DeepSeek V3.2 Speciale DeepSeek Open WeightsStructured | 32.5 | 50.5 | Unknown | — | — | 87.1 | 13.0 | $0.2800/M Cheapest verified | — |
277 | DeepSeek-V4-Pro DeepSeek Ready to UseStructured | 32.4 | 64.6 | 1.6T | — | — | 90.1 | 13.0 | Free Free | — |
335 | MiniMax-M2 MiniMax Open WeightsStructured | 31.6 | 54.6 | 229B | — | — | 77.7 | — | Free Free | — |
408 | GLM-4.5-Air Z.ai Open WeightsStructured | 27.6 | 53.6 | 110B | — | — | 73.3 | — | Free Free | — |