Listings
All listings: Parasail
Every listing for the models on page 2 of 3 of Parasail, including each model's other versions and routes. Back to Parasail
Sold by Parasail
22 listings on this page sold by Parasail.
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 |
|---|---|---|---|---|---|---|---|---|
GPT OSS 120B πΊπΈMaker 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. | 86Β Β±5Coding IQ 85.5, 95% interval 80β91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15 | $0.75 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 Β±11Terse (37 Β±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 πΊπΈMaker 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. | 86Β Β±5Coding IQ 85.5, 95% interval 80β91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10cache read $0.055 | $0.75 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 Β±11Terse (37 Β±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 π¨π³Maker 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. | 98Β Β±7Coding IQ 97.7, 95% interval 90β105. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.12cache read $0.07 | $0.80 | 262K | 36.8Poor value (36.8). Costs more per task than its benchmark results predict. Plenty of data behind it. | 54 Β±9Typical (54 Β±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. | βNot measured |
Qwen3 Next 80B A3B Instruct π¨π³Maker 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. | 92Β Β±11Coding IQ 91.7, 95% interval 81β102. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10cache read $0.07 | $1.10 | 262K | 47.6Fair value (47.6). Costs per task about what its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured |
Mistral Small 3.2 24B Instruct π«π·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. | 113.3From 50 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.09cache read $0.05 | $0.30 | 131K | 58.5Good value (58.5). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured |
Qwen3.5 9B π¨π³Maker 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. | 113.2From 223 benchmarks. Plenty of data behind it. | 86Β Β±8Coding IQ 85.6, 95% interval 77β94. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10 | $0.25 | 262K | 55.4Good value (55.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | βNot measured | 139.5Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API |
Llama 4 Maverick πΊπΈMaker 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. | 78Β Β±9Coding IQ 78.1, 95% interval 69β87. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.35cache read $0.17 | $1.00 | 524K | 42.1Poor value (42.1). Costs more per task than its benchmark results predict. Plenty of data behind it. | 38 Β±9Terse (38 Β±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 πΊπΈMaker 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. | 88Β Β±12Coding IQ 88.2, 95% interval 76β101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.03cache read $0.02 | $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 Β±9Terse (40 Β±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 |
Qwen3 VL 8B Instruct π¨π³Maker 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. | 104.9From 81 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.25cache read $0.12 | $0.75 | 262K | 27.3Poor value (27.3). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured |
BGE M3 π¨π³Maker 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. | 100.9From 136 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.01 | β | 8K | βNot enough data yet | βNot measured | βNot measured |
Gemma 3 4B IT πΊπΈMaker 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. | 95.2From 249 benchmarks. Plenty of data behind it. | 67Β Β±19Coding IQ 66.8, 95% interval 48β86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.04cache read $0.02 | $0.05 | 131K | 38.3Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured |
L3 8B Lunaris V1parasail-sao10kl3-lunaris-8bparasail-sao10kl3-lunaris-8b | 93.9From 6 benchmarks. A fair amount of data behind it. | βNot enough data yet | $0.04 | $0.05 | 8K | βNot enough data yet | βNot measured | βNot measured |
Llama 3.2 3B πΊπΈMaker 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. | 86.5From 255 benchmarks. Plenty of data behind it. | β 76Β Β±14Coding IQ 75.7, 95% interval 62β89. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.05 | $0.33 | 131K | 44.3Poor value (44.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured |
Mistral Nemo Instruct 2407 FP8parasail-mistralaimistral-nemo Β· fp8parasail-mistralaimistral-nemo Β· fp8 | βNot enough data yet | βNot enough data yet | $0.03 | $0.03 | 131K | βNot enough data yet | βNot measured | βNot measured |
Gemma 3 27b It FP8 Dynamicparasail-gemma3-27b-itparasail-gemma3-27b-it | βNot enough data yet | βNot enough data yet | $0.08cache read $0.04 | $0.45 | 131K | βNot enough data yet | βNot measured | βNot measured |
MythoMax L2 13b Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | βNot enough data yet | βNot enough data yet | $0.08 | $0.11 | 4K | βNot enough data yet | βNot measured | βNot measured |
UI TARS 1.5 7Bparasail-ui-tars-1p5-7bparasail-ui-tars-1p5-7b | βNot enough data yet | βNot enough data yet | $0.10 | $0.20 | 128K | βNot enough data yet | βNot measured | βNot measured |
DeepSeek V4 Flash DSpark π¨π³Maker headquartered in CN, according to the company's own website. The citation is on the model page. | βNot enough data yet | βNot enough data yet | $0.14cache read $0.07 | $0.28 | 1M | βNot enough data yet | βNot measured | βNot measured |
Gemma 4 31B It FP8 Dynamicparasail-gemma-4-31b-itparasail-gemma-4-31b-it | βNot enough data yet | βNot enough data yet | $0.15cache read $0.06 | $0.40 | 262K | βNot enough data yet | βNot measured | βNot measured |
GLM 5.3 Flash NVFP4parasail-glm-53-flash-fast Β· nvfp4parasail-glm-53-flash-fast Β· nvfp4 | βNot enough data yet | βNot enough data yet | $0.1875cache read $0.0375 | $0.625 | 1M | βNot enough data yet | βNot measured | βNot measured |
Llama 3.3 70B Instruct FP8 Dynamicparasail-llama-33-70b-fp8 Β· fp8parasail-llama-33-70b-fp8 Β· fp8 | βNot enough data yet | βNot enough data yet | $0.22cache read $0.11 | $0.50 | 131K | βNot enough data yet | βNot measured | βNot measured |
Cydonia 24B V4.1 Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. Newer version | βNot enough data yet | βNot enough data yet | $0.30cache read $0.15 | $0.50 | 131K | 55.4Good value (55.4). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured |
Run by Parasail
43 listings on this page where Parasail 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 |
|---|---|---|---|---|---|---|---|---|---|
GPT OSS 120B πΊπΈMaker 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. | 86Β Β±5Coding IQ 85.5, 95% interval 80β91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10 | $0.75 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 Β±11Terse (37 Β±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 πΊπΈMaker 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. | 86Β Β±5Coding IQ 85.5, 95% interval 80β91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.105cache read $0.0578 | $0.7875 | βNot published by the provider | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 Β±11Terse (37 Β±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 πΊπΈMaker 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. | 86Β Β±5Coding IQ 85.5, 95% interval 80β91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.15 | $0.75 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 Β±11Terse (37 Β±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 πΊπΈMaker 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. | 86Β Β±5Coding IQ 85.5, 95% interval 80β91. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10cache read $0.055 | $0.75 | 131K | 58.4Good value (58.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | 37 Β±11Terse (37 Β±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 π¨π³Maker 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. | 98Β Β±7Coding IQ 97.7, 95% interval 90β105. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.12 | $0.80 | 262K | 36.8Poor value (36.8). Costs more per task than its benchmark results predict. Plenty of data behind it. | 54 Β±9Typical (54 Β±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. | βNot measured | |
Qwen3 Coder Next π¨π³Maker 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. | 98Β Β±7Coding IQ 97.7, 95% interval 90β105. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.126cache read $0.0735 | $0.84 | βNot published by the provider | 36.8Poor value (36.8). Costs more per task than its benchmark results predict. Plenty of data behind it. | 54 Β±9Typical (54 Β±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. | βNot measured | |
Qwen3 Coder Next π¨π³Maker 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. | 98Β Β±7Coding IQ 97.7, 95% interval 90β105. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.12cache read $0.07 | $0.80 | 262K | 36.8Poor value (36.8). Costs more per task than its benchmark results predict. Plenty of data behind it. | 54 Β±9Typical (54 Β±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. | βNot measured | |
Qwen3 Next 80B A3B Instruct π¨π³Maker 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. | 92Β Β±11Coding IQ 91.7, 95% interval 81β102. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10 | $1.10 | 262K | 47.6Fair value (47.6). Costs per task about what its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured | |
Qwen3 Next 80B A3B Instruct π¨π³Maker 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. | 92Β Β±11Coding IQ 91.7, 95% interval 81β102. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.105cache read $0.0735 | $1.155 | βNot published by the provider | 47.6Fair value (47.6). Costs per task about what its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured | |
Qwen3 Next 80B A3B Instruct π¨π³Maker 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. | 92Β Β±11Coding IQ 91.7, 95% interval 81β102. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10cache read $0.07 | $1.10 | 262K | 47.6Fair value (47.6). Costs per task about what its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured | |
Mistral Small 3.2 24B Instruct π«π·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. | 113.3From 50 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.09cache read $0.05 | $0.30 | 131K | 58.5Good value (58.5). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Qwen3.5 9B π¨π³Maker 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. | 113.2From 223 benchmarks. Plenty of data behind it. | 86Β Β±8Coding IQ 85.6, 95% interval 77β94. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10 | $0.25 | 262K | 55.4Good value (55.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | βNot measured | 139.5Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Qwen3.5 9B π¨π³Maker 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. | 113.2From 223 benchmarks. Plenty of data behind it. | 86Β Β±8Coding IQ 85.6, 95% interval 77β94. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.105 | $0.2625 | βNot published by the provider | 55.4Good value (55.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | βNot measured | 139.5Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Qwen3.5 9B π¨π³Maker 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. | 113.2From 223 benchmarks. Plenty of data behind it. | 86Β Β±8Coding IQ 85.6, 95% interval 77β94. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.10 | $0.25 | 262K | 55.4Good value (55.4). Cheaper per task than its benchmark results predict. Plenty of data behind it. | βNot measured | 139.5Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Llama 4 Maverick πΊπΈMaker 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. | 78Β Β±9Coding IQ 78.1, 95% interval 69β87. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.35 | $1.00 | 524K | 42.1Poor value (42.1). Costs more per task than its benchmark results predict. Plenty of data behind it. | 38 Β±9Terse (38 Β±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 | |
Llama 4 Maverick πΊπΈMaker 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. | 78Β Β±9Coding IQ 78.1, 95% interval 69β87. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.3675cache read $0.1785 | $1.05 | βNot published by the provider | 42.1Poor value (42.1). Costs more per task than its benchmark results predict. Plenty of data behind it. | 38 Β±9Terse (38 Β±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 | |
Llama 4 Maverick πΊπΈMaker 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. | 78Β Β±9Coding IQ 78.1, 95% interval 69β87. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.35cache read $0.17 | $1.00 | 524K | 42.1Poor value (42.1). Costs more per task than its benchmark results predict. Plenty of data behind it. | 38 Β±9Terse (38 Β±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 πΊπΈMaker 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. | 88Β Β±12Coding IQ 88.2, 95% interval 76β101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.03 | $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 Β±9Terse (40 Β±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 πΊπΈMaker 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. | 88Β Β±12Coding IQ 88.2, 95% interval 76β101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.0315cache read $0.021 | $0.1575 | βNot published by the provider | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 Β±9Terse (40 Β±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 πΊπΈMaker 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. | 88Β Β±12Coding IQ 88.2, 95% interval 76β101. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.03cache read $0.02 | $0.15 | 131K | 48.9Fair value (48.9). Costs per task about what its benchmark results predict. Plenty of data behind it. | 40 Β±9Terse (40 Β±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 | |
Qwen2.5 VL 72B Instruct π¨π³Maker 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. | 108.8From 97 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.80 | $1.00 | 128K | 41.2Poor value (41.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Qwen2.5 VL 72B Instruct π¨π³Maker 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. | 108.8From 97 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.80cache read $0.40 | $1.00 | 128K | 41.2Poor value (41.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Qwen2.5 VL 72B Instruct π¨π³Maker 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. | 108.8From 97 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.70 | $0.70 | 33K | 41.2Poor value (41.2). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Gemma 3 27B IT πΊπΈMaker 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. | 108.5From 365 benchmarks. Plenty of data behind it. | 73Β Β±8Coding IQ 72.7, 95% interval 65β81. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.08 | $0.45 | 131K | 46.2Poor value (46.2). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | 127.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Gemma 3 27B IT πΊπΈMaker 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. | 108.5From 365 benchmarks. Plenty of data behind it. | 73Β Β±8Coding IQ 72.7, 95% interval 65β81. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.08cache read $0.04 | $0.45 | 131K | 46.2Poor value (46.2). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | 127.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Gemma 3 27B IT πΊπΈMaker 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. | 108.5From 365 benchmarks. Plenty of data behind it. | 73Β Β±8Coding IQ 72.7, 95% interval 65β81. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.30 | $0.50 | 128K | 46.2Poor value (46.2). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | 127.6Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Llama 3.3 70B Instruct πΊπΈMaker 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. | 108.1From 278 benchmarks. Plenty of data behind it. | 77Β Β±9Coding IQ 77.1, 95% interval 68β86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.22 | $0.50 | 131K | 31.6Poor value (31.6). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | 126.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Llama 3.3 70B Instruct πΊπΈMaker 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. | 108.1From 278 benchmarks. Plenty of data behind it. | 77Β Β±9Coding IQ 77.1, 95% interval 68β86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.231cache read $0.1155 | $0.525 | βNot published by the provider | 31.6Poor value (31.6). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | 126.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Llama 3.3 70B Instruct πΊπΈMaker 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. | 108.1From 278 benchmarks. Plenty of data behind it. | 77Β Β±9Coding IQ 77.1, 95% interval 68β86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.22cache read $0.11 | $0.50 | 131K | 31.6Poor value (31.6). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | 126.2Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Qwen3 VL 8B Instruct π¨π³Maker 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. | 104.9From 81 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.25 | $0.75 | 262K | 27.3Poor value (27.3). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured | |
Qwen3 VL 8B Instruct π¨π³Maker 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. | 104.9From 81 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.25cache read $0.12 | $0.75 | 262K | 27.3Poor value (27.3). Costs more per task than its benchmark results predict. A fair amount of data behind it. | βNot measured | βNot measured | |
Mistral Nemo π«π·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. | 98.1From 85 benchmarks. Plenty of data behind it. | βNot enough data yet | $0.03 | $0.03 | 131K | 52.2Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | 113.9Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API | |
Gemma 3 4B IT πΊπΈMaker 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. | 95.2From 249 benchmarks. Plenty of data behind it. | 67Β Β±19Coding IQ 66.8, 95% interval 48β86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.04 | $0.05 | 131K | 38.3Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Gemma 3 4B IT πΊπΈMaker 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. | 95.2From 249 benchmarks. Plenty of data behind it. | 67Β Β±19Coding IQ 66.8, 95% interval 48β86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.04cache read $0.02 | $0.05 | 131K | 38.3Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
93.9From 6 benchmarks. A fair amount of data behind it. | βNot enough data yet | $0.04 | $0.05 | 8K | βNot enough data yet | βNot measured | βNot measured | ||
Llama 3.2 3B πΊπΈMaker 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. | 86.5From 255 benchmarks. Plenty of data behind it. | β 76Β Β±14Coding IQ 75.7, 95% interval 62β89. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.05 | $0.33 | 131K | 44.3Poor value (44.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Llama 3.2 3B πΊπΈMaker 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. | 86.5From 255 benchmarks. Plenty of data behind it. | β 76Β Β±14Coding IQ 75.7, 95% interval 62β89. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.0525 | $0.3465 | βNot published by the provider | 44.3Poor value (44.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
Llama 3.2 3B πΊπΈMaker 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. | 86.5From 255 benchmarks. Plenty of data behind it. | β 76Β Β±14Coding IQ 75.7, 95% interval 62β89. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ. | $0.05 | $0.33 | 131K | 44.3Poor value (44.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured | |
MythoMax L2 13b Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | βNot enough data yet | βNot enough data yet | $0.08 | $0.11 | 4K | βNot enough data yet | βNot measured | βNot measured | |
MythoMax L2 13b Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. | βNot enough data yet | βNot enough data yet | $0.08 | $0.11 | 4K | βNot enough data yet | βNot measured | βNot measured | |
| βNot enough data yet | βNot enough data yet | $0.10 | $0.20 | 128K | βNot enough data yet | βNot measured | βNot measured | ||
UI TARS 1.5 7Bbf16bf16 | βNot enough data yet | βNot enough data yet | $0.10 | $0.20 | 128K | βNot enough data yet | βNot measured | βNot measured | |
Cydonia 24B V4.1 Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo. Newer version | βNot enough data yet | βNot enough data yet | $0.30cache read $0.15 | $0.50 | 131K | 55.4Good value (55.4). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate. | βNot measured | βNot measured |
Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)