Text generation and understanding
20
Models
21
Benchmarks
9
Providers
18
Open Weight
19
Deployable Now
18 self-host
20
Structured benchmark-backed rows
0
Signal-backed rows without full normalized tables
0
Benchmark-expected rows still catching up
| # | Model | Economic Footprint | Quality | Params | MMLU | BigCodeBench | LiveBench Language | TerminalBench 2.0 | $/M In | Speed | Open |
|---|---|---|---|---|---|---|---|---|---|---|---|
130 | Command R+ Cohere Open Weights |
| 53.6 |
| 63.6 |
| 104B |
| 75.7 |
| 41.9 |
| — |
| — |
$2.50/M Cheapest verified |
| — |
162 | Mistral Large 2 Mistral AI Structured | 52.7 | 57.9 | 123B | — | — | 2.3 | 38.7 | $2.00/M Cheapest verified | — |
92 | llama-3-70b-instruct Meta Open WeightsStructured | 51.0 | 63.4 | 70B | 82.0 | 54.5 | 31.0 | — | Free Free | — |
99 | llama-3.1-405b-instruct Meta Open WeightsStructured | 49.9 | 61.7 | 405B | 87.3 | — | — | — | Free Cheapest verified | — |
15 | QwQ 32B Qwen Open WeightsStructured | 48.0 | 64.1 | 32B | 76.4 | — | 73.8 | — | $0.1500/M Cheapest verified | — |
52 | Llama 3.1 Nemotron 70B Instruct NVIDIA Open WeightsStructured | 47.8 | 51.4 | 70B | — | 48.2 | 17.2 | — | $1.20/M Cheapest verified | — |
72 | Llama 3.3 70B Meta Open WeightsStructured | 47.5 | 68.3 | 70B | 86.0 | — | — | — | $0.5400/M Cheapest verified | — |
85 | Qwen2.5-72B Alibaba Open WeightsStructured | 47.4 | 67.4 | 72B | 86.1 | — | — | — | $0.6000/M Cheapest verified | — |
30 | DeepSeek-V3 DeepSeek Open WeightsStructured | 47.2 | 79.6 | 685B | 87.1 | 62.2 | 48.1 | — | $0.2800/M Cheapest verified | 145 tok/s |
40 | DeepSeek-R1 DeepSeek Self-HostStructured | 46.8 | 80.4 | 671B | 90.8 | — | 65.5 | — | $0.2800/M Cheapest verified | — |
63 | Gemma 3 27B | 46.4 | 63.5 | 27B | — | — | — | — | Free Cheapest verified | — |
42 | Llama 4 Maverick Meta Open WeightsStructured | 45.4 | 72.9 | 400B | 80.9 | 61.4 | — | 39.8 | Free Cheapest verified | 88 tok/s |
136 | Nemotron 70B NVIDIA Open WeightsStructured | 45.3 | 65.5 | 70B | — | — | — | — | Free Cheapest verified | — |
39 | Qwen3-235B Alibaba Self-HostStructured | 43.6 | 79.6 | 235B | — | — | — | 45.3 | $0.8000/M Cheapest verified | 72 tok/s |
95 | grok-4 xAI SubscriptionStructured | 42.6 | 67.0 | Undisclosed | 86.6 | — | — | 27.2 | $3.00/M Cheapest verified | — |
203 | mixtral-8x7b-instruct-v0.1 Mistral AI Open WeightsStructured | 38.7 | 43.7 | 56B | — | — | 36.5 | — | Free Free | — |
128 | DeepSeek-V3.2 DeepSeek Open WeightsStructured | 34.1 | 72.6 | Unknown | — | — | 48.1 | 39.6 | $0.2800/M Cheapest verified | — |
233 | Qwen2.5-7B-Instruct Qwen Open WeightsStructured | 33.5 | 43.4 | 7B | — | 46.1 | — | — | Free Free | — |
207 | mistral-7b-instruct-v0.2 Mistral AI Open WeightsStructured | 32.5 | 37.7 | 7B | — | — | 3.2 | — | Free Free | — |
221 | Qwen3-0.6B Qwen Self-HostStructured | 30.3 | 45.5 | 600M | — | — | — | — | Free Free | — |