Listings
All listings: Amazon Web Services
Every listing for the models on page 3 of 9 of Amazon Web Services, including each model's other versions and routes. Back to Amazon Web Services
Sold by Amazon Web Services
60 listings on this page sold by Amazon Web Services.
Select a score for details.
About these columns
- llminfo IQ
- llminfo IQ (capability index). How llminfo IQ is computed
- Coding IQ
- llminfo Coding IQ (coding capability, its own scale). How Coding IQ is computed
- Value
- Value for money. 50 means as expected for its results. Above 50, cheaper per task. How Value for money works
- Yap
- Tokens used per coding task vs. models with the same results. Includes agents that write code. Chattiness on maths doesn't predict it. How the Yap score works
- Epoch Capabilities Index
- Epoch Capabilities Index benchmark details
| Model | llminfo IQ | Coding IQ | Input $/1M | Output $/1M | Max Input Tokens | Value | Yap | Epoch Capabilities Index |
|---|---|---|---|---|---|---|---|---|
DeepSeek V3.1 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 119.6From 122 benchmarks. Plenty of data behind it. | 94 ±8Coding IQ 94.3, 95% interval 86–103. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.58batch $0.29 | $1.68batch $0.84 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 140.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Devstral 2 123b Instruct 2512 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 119.5From 25 benchmarks. Plenty of data behind it. | 97 ±9Coding IQ 97.1, 95% interval 88–106. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.40batch $0.20 | $2.00batch $1.00 | 256K | 50.2Fair value (50.2). Costs per task about what its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | —Not published by the provider | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | —Not published by the provider | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00 | $15.00 | —Not published by the provider | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | 200K | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | 200K | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | 200K | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | 200K | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.60batch $1.80 | $18.00batch $9.00 | 200K | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00cache read $0.30 | $15.00cache write $3.75 | 200K | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00 | $15.00 | —Not published by the provider | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3.5 Sonnet 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 119.3From 546 benchmarks. Plenty of data behind it. | 89 ±9Coding IQ 89.4, 95% interval 81–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $3.00 | $15.00 | —Not published by the provider | 27.0Poor value (27.0). Costs more per task than its benchmark results predict. Plenty of data behind it. | —Not measured | 130Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Qwen3 Coder 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 119.2From 84 benchmarks. Plenty of data behind it. | 93 ±6Coding IQ 92.6, 95% interval 87–98. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.45batch $0.225 | $1.80batch $0.90 | 262K | 34.9Poor value (34.9). Costs more per task than its benchmark results predict. Plenty of data behind it. | 39 ±9Terse (39 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | —Not measured |
MiniMax M2 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.9From 111 benchmarks. Plenty of data behind it. | 95 ±7Coding IQ 95.4, 95% interval 88–103. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.30batch $0.15 | $1.20batch $0.60 | 1M | 57.8Good value (57.8). Cheaper per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured |
Mistral Large 3 675b Instruct 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.9From 35 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.50batch $0.25 | $1.50batch $0.75 | 256K | 53.7Good value (53.7). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15batch $0.09 | $0.60batch $0.36 | 128K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15batch $0.075 | $0.60batch $0.30 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.18 | $0.72 | 128K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Qwen3 Coder Next 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 27 benchmarks. Plenty of data behind it. | 98 ±7Coding IQ 97.7, 95% interval 90–105. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.50batch $0.25 | $1.20batch $0.60 | 256K | 36.8Poor value (36.8). Costs more per task than its benchmark results predict. Plenty of data behind it. | 54 ±9Typical (54 ±9). Generates about as many tokens as models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | —Not measured |
Qwen3 Next 80B A3B Instruct 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.1From 154 benchmarks. Plenty of data behind it. | 92 ±11Coding IQ 91.7, 95% interval 81–102. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.14batch $0.07 | $1.20batch $0.60 | 256K | 47.6Fair value (47.6). Costs per task about what its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured |
Qwen3 Coder 30B A3B Instruct 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 116.2From 18 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.15batch $0.075 | $0.60batch $0.30 | 262K | 54.3Good value (54.3). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Qwen3 Coder 30B A3B Instruct 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 116.2From 18 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.15batch $0.075 | $0.60batch $0.30 | 256K | 54.3Good value (54.3). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
GLM 4.7 Flash 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. Newer version | 116.1From 186 benchmarks. Plenty of data behind it. | 83 ±10Coding IQ 82.8, 95% interval 73–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07batch $0.035 | $0.40batch $0.20 | 203K | 44.1Poor value (44.1). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured |
Qwen3 32B 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 115.6From 234 benchmarks. Plenty of data behind it. | 85 ±9Coding IQ 84.8, 95% interval 76–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15batch $0.075 | $0.60batch $0.30 | 33K | 35.8Poor value (35.8). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | 136.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Qwen3 32B 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 115.6From 234 benchmarks. Plenty of data behind it. | 85 ±9Coding IQ 84.8, 95% interval 76–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15batch $0.075 | $0.60batch $0.30 | 32K | 35.8Poor value (35.8). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | 136.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Nova Premier 1.0 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | ≈ 115From 13 benchmarks. Little data behind it, so treat it as a rough estimate. | —Not enough data yet | $2.50cache read $0.625batch $1.25 | $12.50batch $6.25 | 1M | 43.5Poor value (43.5). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova Premier 1.0 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | ≈ 115From 13 benchmarks. Little data behind it, so treat it as a rough estimate. | —Not enough data yet | $2.50 | $12.50 | —Not published by the provider | 43.5Poor value (43.5). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova Premier 1.0 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | ≈ 115From 13 benchmarks. Little data behind it, so treat it as a rough estimate. | —Not enough data yet | $2.50cache read $0.625 | $12.50cache write $2.50 | 1M | 43.5Poor value (43.5). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova Premier 1.0 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | ≈ 115From 13 benchmarks. Little data behind it, so treat it as a rough estimate. | —Not enough data yet | $2.50 | $12.50 | 1M | 43.5Poor value (43.5). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
DeepSeek V3 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Newer version | 114.5From 492 benchmarks. Plenty of data behind it. | 85 ±7Coding IQ 85.3, 95% interval 79–92. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.7017 | $2.0324 | 128K | 43.6Poor value (43.6). Costs more per task than its benchmark results predict. Plenty of data behind it. | 47 ±9Typical (47 ±9). Generates about as many tokens as models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 135.3Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.30cache read $0.075batch $0.15 | $2.50batch $1.25 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | —Not published by the provider | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.30cache read $0.075batch $0.15 | $2.50batch $1.25 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.33cache read $0.0825 | $2.75 | —Not published by the provider | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.396cache read $0.099 | $3.311 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.363cache read $0.0908 | $2.992 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.396cache read $0.099batch $0.198 | $3.311cache write $0.396batch $1.6555 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.33cache read $0.0825batch $0.165 | $2.75batch $1.375 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Llama 3.1 405B Instruct 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 111.8From 123 benchmarks. Plenty of data behind it. | 84 ±9Coding IQ 84.3, 95% interval 76–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $2.40 | $2.40 | 128K | 41.9Poor value (41.9). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Llama 3.1 405B Instruct 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 111.8From 123 benchmarks. Plenty of data behind it. | 84 ±9Coding IQ 84.3, 95% interval 76–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $2.40 | $2.40 | 128K | 41.9Poor value (41.9). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Llama 4 Maverick 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.8From 350 benchmarks. Plenty of data behind it. | 78 ±9Coding IQ 78.1, 95% interval 69–87. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.24batch $0.12 | $0.97batch $0.485 | 1M | 42.1Poor value (42.1). Costs more per task than its benchmark results predict. Plenty of data behind it. | 38 ±9Terse (38 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 132.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07batch $0.042 | $0.30batch $0.18 | 128K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07batch $0.035 | $0.30batch $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.084 | $0.36 | 128K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3 Opus 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 111.1From 307 benchmarks. Plenty of data behind it. | 80 ±10Coding IQ 79.6, 95% interval 70–89. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $15.00 | $75.00 | —Not published by the provider | 38.0Poor value (38.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | 125.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3 Opus 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 111.1From 307 benchmarks. Plenty of data behind it. | 80 ±10Coding IQ 79.6, 95% interval 70–89. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $15.00cache read $1.50 | $75.00cache write $18.75 | 200K | 38.0Poor value (38.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | 125.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3 Opus 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 111.1From 307 benchmarks. Plenty of data behind it. | 80 ±10Coding IQ 79.6, 95% interval 70–89. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $15.00cache read $1.50 | $75.00cache write $18.75 | 200K | 38.0Poor value (38.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | 125.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3 Opus 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 111.1From 307 benchmarks. Plenty of data behind it. | 80 ±10Coding IQ 79.6, 95% interval 70–89. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $15.00cache read $1.50 | $75.00cache write $18.75 | 200K | 38.0Poor value (38.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | 125.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Claude 3 Opus 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Newer version | 111.1From 307 benchmarks. Plenty of data behind it. | 80 ±10Coding IQ 79.6, 95% interval 70–89. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $15.00 | $75.00 | —Not published by the provider | 38.0Poor value (38.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | 125.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Embed Multilingual V3.0 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | ≈ 111From 29 benchmarks. Little data behind it, so treat it as a rough estimate. | —Not enough data yet | $0.10 | — | 512 | —Not enough data yet | —Not measured | —Not measured |
Embed Multilingual V3.0 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | ≈ 111From 29 benchmarks. Little data behind it, so treat it as a rough estimate. | —Not enough data yet | — | — | —Not published by the provider | —Not enough data yet | —Not measured | —Not measured |
Ministral 3 8b Instruct 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 110.9From 14 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.15batch $0.07 | $0.15batch $0.07 | 128K | —Not enough data yet | —Not measured | —Not measured |
Ministral 3 14B Instruct 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 110.8From 15 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.20batch $0.10 | $0.20batch $0.10 | 128K | —Not enough data yet | —Not measured | —Not measured |
Nemotron Nano 9b V2 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 110.6From 19 benchmarks. Plenty of data behind it. | —Not enough data yet | $0.06batch $0.03 | $0.23batch $0.12 | 128K | 52.1Fair value (52.1). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nemotron Nano 9b V2 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 110.6From 19 benchmarks. Plenty of data behind it. | —Not enough data yet | $0.06 | $0.23 | —Not published by the provider | 52.1Fair value (52.1). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Nemotron Nano 9b V2 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 110.6From 19 benchmarks. Plenty of data behind it. | —Not enough data yet | $0.072 | $0.276 | 128K | 52.1Fair value (52.1). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured |
Run by Amazon Web Services
28 listings on this page where Amazon Web Services runs the model. Another provider sells these listings.
Select a score for details.
About these columns
- llminfo IQ
- llminfo IQ (capability index). How llminfo IQ is computed
- Coding IQ
- llminfo Coding IQ (coding capability, its own scale). How Coding IQ is computed
- Value
- Value for money. 50 means as expected for its results. Above 50, cheaper per task. How Value for money works
- Yap
- Tokens used per coding task vs. models with the same results. Includes agents that write code. Chattiness on maths doesn't predict it. How the Yap score works
- Epoch Capabilities Index
- Epoch Capabilities Index benchmark details
| Model | Provider | llminfo IQ | Coding IQ | Input $/1M | Output $/1M | Max Input Tokens | Value | Yap | Epoch Capabilities Index |
|---|---|---|---|---|---|---|---|---|---|
Devstral 2 123b Instruct 2512 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 119.5From 25 benchmarks. Plenty of data behind it. | 97 ±9Coding IQ 97.1, 95% interval 88–106. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 256K | 50.2Fair value (50.2). Costs per task about what its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured | |
MiniMax M2 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.9From 111 benchmarks. Plenty of data behind it. | 95 ±7Coding IQ 95.4, 95% interval 88–103. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 400K | 57.8Good value (57.8). Cheaper per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured | |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15 | $0.60 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15 | $0.60 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 120B 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 424 benchmarks. Plenty of data behind it. | 86 ±5Coding IQ 85.5, 95% interval 80–91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15 | $0.60 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 ±11Terse (37 ±11). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 140.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Qwen3 Coder Next 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 118.2From 27 benchmarks. Plenty of data behind it. | 98 ±7Coding IQ 97.7, 95% interval 90–105. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.78 | $1.86 | 262K | 36.8Poor value (36.8). Costs more per task than its benchmark results predict. Plenty of data behind it. | 54 ±9Typical (54 ±9). Generates about as many tokens as models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | —Not measured | |
Qwen3 Coder 30B A3B Instruct 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 116.2From 18 benchmarks. A fair amount of data behind it. | —Not enough data yet | — | — | 262K | 54.3Good value (54.3). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Qwen3 Coder 30B A3B Instruct 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 116.2From 18 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.15 | $0.60 | 262K | 54.3Good value (54.3). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Qwen3 Coder 30B A3B Instruct 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 116.2From 18 benchmarks. A fair amount of data behind it. | —Not enough data yet | $0.15 | $0.60 | 262K | 54.3Good value (54.3). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
GLM 4.7 Flash 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. Newer version | 116.1From 186 benchmarks. Plenty of data behind it. | 83 ±10Coding IQ 82.8, 95% interval 73–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 203K | 44.1Poor value (44.1). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured | |
GLM 4.7 Flash 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. Newer version | 116.1From 186 benchmarks. Plenty of data behind it. | 83 ±10Coding IQ 82.8, 95% interval 73–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07 | $0.40 | 203K | 44.1Poor value (44.1). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured | |
GLM 4.7 Flash 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. Newer version | 116.1From 186 benchmarks. Plenty of data behind it. | 83 ±10Coding IQ 82.8, 95% interval 73–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07 | $0.40 | 203K | 44.1Poor value (44.1). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | —Not measured | |
Ministral 14B 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 115.6From 33 benchmarks. A fair amount of data behind it. | —Not enough data yet | — | — | 256K | —Not enough data yet | —Not measured | —Not measured | |
Qwen3 32B 🇨🇳Maker headquartered in CN, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 115.6From 234 benchmarks. Plenty of data behind it. | 85 ±9Coding IQ 84.8, 95% interval 76–93. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 16K | 35.8Poor value (35.8). Costs more per task than its benchmark results predict. A fair amount of data behind it. | —Not measured | 136.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | — | — | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | —Not published by the provider | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Pixtral Large 🇫🇷Maker headquartered in FR, according to the company's own website. The citation is on the model page. | 114.2From 15 benchmarks. Plenty of data behind it. | —Not enough data yet | $2.00 | $6.00 | 128K | 44.0Poor value (44.0). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.30 | $2.50 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Nova 2 Lite 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. | 112.6From 78 benchmarks. Plenty of data behind it. | 94 ±41Coding IQ 94.4, 95% interval 53–136. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.30 | $2.50 | 1M | 38.2Poor value (38.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | —Not measured | —Not measured | |
Llama 4 Maverick 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.8From 350 benchmarks. Plenty of data behind it. | 78 ±9Coding IQ 78.1, 95% interval 69–87. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.24 | $0.97 | —Not published by the provider | 42.1Poor value (42.1). Costs more per task than its benchmark results predict. Plenty of data behind it. | 38 ±9Terse (38 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 132.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | — | — | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07 | $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07 | $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
GPT OSS 20b 🇺🇸Maker headquartered in US, according to the company's own website. The citation is on the model page. Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | 111.3From 251 benchmarks. Plenty of data behind it. | 88 ±12Coding IQ 88.2, 95% interval 76–101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.07 | $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 ±9Terse (40 ±9). Generates fewer tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. | 134.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)